Chapter 9 Part Two: Results and Discussion

Analysis 2: A second set of key words from the PMC was then computed – resulting in Analysis 2 – this time using the BEC as the reference point. The key words formed by this second analysis were those which occurred significantly more frequently in published Business English materials than in the Business English of the BEC. Analysis of these key words, therefore, was able to show how the lexis of published Business English materials differs from the lexis of the real-life Business English found in the BEC. This process of analysis is summarised in the diagram below:

PMC -> (BNC) PMC -> (BEC)

Published Business English Published Business English

(General English reference corpus) (Real Business reference corpus)

Key Words Key Words

Analysis Analysis

Comparison of PMC key words: 1) computed using BNC as reference

2) computed using BEC as reference

Fig. 54 The two-way process of key word analysis of PMC lexis

2. Fewer words: The same analytical procedures – semantic prosody, colligation and clusters[196] – carried out on the fifty words in the BEC were carried out on words from the PMC. However, in this case only five words were chosen for comparative analysis. These words were manager, customer, product, market and business. All of these words were in the ten most ‘key’ key words of the PMC (using the BNC as a reference corpus).

Four questions were asked of the lexis of the PMC:

1. How do the PMC key words define the lexical world of business and how does this definition compare to that shown in the BEC?

2. What semantic prosodies were found in the PMC and do they match or differ from those found in the BEC?

3. What colligational and grammar/meaning patterns were identified in the PMC and how do they compare to those found in the BEC?

4. How do the clusters found in the PMC compare to those in the BEC?

Each of these questions will now be discussed in turn, leading to a justification of Hypothesis 2, and a final analysis and answering of the main research question.

9.4.1 How do the PMC key words define the lexical world of business and how does this definition compare to that shown in the BEC?

The analysis of key words from the PMC in this section is, therefore, carried out in two parts. Firstly, the PMC key words computed from statistical comparison to the BNC are analysed – Analysis 1 shown below – and secondly, the PMC key words gained from statistical comparison to the BEC are analysed – Analysis 2. Final conclusions are then drawn from both sets of analyses.

9.4.1.1 Analysis 1:

Key word analysis of the PMC (with the BNC as the reference corpus)[197]

The PMC corpus was lemmatised using the modified lemma list of Someya (1998), and was statistically compared to the lemmatised BNC Sampler corpus using the Log Likelihood statistic (Dunning 1993), with a p value of p = 0.000001. A key word list was generated and then manually edited to sift out names of people, products, places, currencies and companies. The full key word list can be found on the CD ROM inside the back page of this thesis. The top 100 key words are printed below:

TABLE LXIV: THE 100 MOST ‘KEY’ KEY WORDS OF THE PMC (BNC REFERENCE CORPUS)

N	WORD	FREQ.	PMC.LST %	FREQ.	BNC.LST %	KEYNESS Log L.
1	COMPANY	2 045	0.34	782	0.04	3 054.4
2	MARKET	1 478	0.25	831	0.04	1 739.6
3	OUR	2 539	0.43	2 577	0.13	1 687.7
4	SALE	968	0.16	343	0.02	1 501.7
5	PRODUCT	1 006	0.17	412	0.02	1 447.1
6	BUSINESS	1 085	0.18	542	0.03	1 384.9
7	MANAGER	811	0.14	317	0.02	1 196.3
8	PLEASE	1 079	0.18	659	0.03	1 193.0
9	WE	5 837	0.98	10 822	0.55	1 192.7
10	OK	483	0.08	38		1 158.5
11	PRICE	949	0.16	586	0.03	1 040.2
12	YOUR	2 949	0.50	4 670	0.24	912.2
13	CUSTOMER	528	0.09	147		912.2
14	BANK	694	0.12	379	0.02	833.5
15	EMPLOYEE	411	0.07	94		764.6
16	FAX	288	0.05	32		649.9
17	YOU	10 587	1.78	26 331	1.34	594.8
18	MEETING	697	0.12	575	0.03	587.8
19	DEPARTMENT	456	0.08	232	0.01	574.8
20	ORDER	747	0.13	681	0.03	564.9
21	CREDIT	342	0.06	110		555.2
22	LTD	332	0.06	103		547.8
23	AFRAID	358	0.06	145		517.9
24	SINCERELY	204	0.03	11		514.7
25	JOB	755	0.13	752	0.04	513.2
26	DIRECTOR	464	0.08	289	0.01	504.9
27	OFFICE	568	0.10	461	0.02	487.0
28	PRODUCTION	362	0.06	167		485.6
29	THANK	877	0.15	1 015	0.05	484.9
30	INVOICE	203	0.03	17		482.1
31	DISCOUNT	225	0.04	36		466.6
32	NEW	1 332	0.22	1 995	0.10	465.3
33	UM	168	0.03	4		454.7
34	BRAND	270	0.05	82		449.6
35	PER	619	0.10	585	0.03	448.7
36	TELEX	167	0.03	6		438.7
37	DELIVERY	237	0.04	56		435.8
38	TO	17 365	2.93	47 851	2.44	421.0
39	PERSONNEL	219	0.04	50		407.6
40	I’M	1 759	0.30	3 136	0.16	401.3
41	ADVERTISE	274	0.05	110		398.3
42	ENCLOSE	186	0.03	27		395.4
43	SELL	487	0.08	419	0.02	392.7
44	MANAGEMENT	398	0.07	279	0.01	392.2
45	PAYMENT	266	0.04	115		370.9
46	INTEREST	716	0.12	865	0.04	370.4
47	FLIGHT	225	0.04	72		366.1
48	AGREE	482	0.08	443	0.02	361.0
49	YOURS	302	0.05	166		360.9
50	REF	134	0.02	5		350.9
51	COST	632	0.11	747	0.04	338.3
52	SUPPLIER	184	0.03	44		336.9
53	COMPANY’S	181	0.03	45		326.7
54	HELLO	303	0.05	192		325.2
55	CORPORATE	164	0.03	33		318.3
56	OFFER	481	0.08	491	0.03	316.8
57	MEET	456	0.08	457	0.02	307.5
58	TELEPHONE	270	0.05	159		306.8
59	I’D	634	0.11	807	0.04	301.1
60	INTERVIEW	197	0.03	78		288.5
61	EXECUTIVE	200	0.03	86		279.7
62	INTERNATIONAL	328	0.06	269	0.01	278.2
63	GOOD	1 456	0.25	2 754	0.14	278.0
64	MILLION	443	0.07	473	0.02	274.8
65	CAN	2 545	0.43	5 606	0.29	273.8
66	CONSIGNMENT	91	0.02	0		265.7
67	CENT	375	0.06	364	0.02	263.1
68	DATE	386	0.07	389	0.02	258.4
69	UK	288	0.05	224	0.01	257.6
70	EXPORT	173	0.03	67		256.3
71	PRESENTATION	168	0.03	62		255.2
72	US	1 186	0.20	2 165	0.11	252.5
73	ACCOUNT	488	0.08	593	0.03	250.1
74	HOW	1 385	0.23	2 666	0.14	249.9
75	DEAR	360	0.06	353	0.02	249.5
76	WORK	1 519	0.26	3 009	0.15	248.7
77	SALARY	126	0.02	25		245.6
78	SORRY	447	0.08	537	0.03	233.3
79	SURE	501	0.08	652	0.03	229.0
80	CONTRACT	242	0.04	183		222.4
81	STAFF	437	0.07	535	0.03	221.3
82	INCREASE	452	0.08	566	0.03	220.6
83	FAITHFULLY	79	0.01	1		220.5
84	WILL	2 243	0.38	5 038	0.26	220.3
85	FOR	6 062	1.02	15 996	0.82	217.5
86	SHARE	544	0.09	762	0.04	217.4
87	MACHINE	293	0.05	274	0.01	215.0
88	MR	1 041	0.18	1 920	0.10	214.8
89	MANUFACTURE	199	0.03	126		213.7
90	FINANCE	192	0.03	117		212.5
91	FIRM	286	0.05	265	0.01	212.2
92	CLOTHES	72	0.01	0		210.2
93	YEAR	1 532	0.26	3 184	0.16	210.1
94	PERFORMANCE	229	0.04	175		208.4
95	MANUFACTURER	138	0.02	53		205.4
96	OPTION	195	0.03	128		203.3
97	CASH	191	0.03	124		200.9
98	PROBLEM	697	0.12	1 136	0.06	200.9
99	GOODBYE	103	0.02	21		199.1
100	CLIENT	191	0.03	126		198.3

A first look reveals the presence of a large number (47 – shown above shaded in grey) of business-related words, which indicates that business lexis features high in the PMC, i.e. it occurs more often than would be expected in general English. Also present are several words that would not look out of place in business, for example, fax, flight, offer, meet, presentation and increase. However, there is also a third prominent group of words related to interpersonal contact, obviously in both spoken and written form, for example, please, afraid, sincerely, thank, enclose, agree, hello, meet, dear, sorry, goodbye. This very cursory look at the most ‘key’ key words shows a trend that will be returned to regularly in this part of the chapter – business is seen in the materials very much in a personal context. Relationships between people are used to generate the business situations presented in the books, in the same way that novelists create a character that their readers can identify with. A further category of key words – personal/possessive pronouns – confirms this – our, we, you, I’m, I’d – all lexis commonly used in reference to people.[198] This may well be an unavoidable compromise in writing to make the materials more interesting, but at the same time it exposes students to lexis that is often the very opposite of what was discovered about business lexis through analysis of the BEC. These initial comments need more explanation, and the following section will do just that.

In order to further analyse the key words of the PMC, they were categorised by word class using the system of Ljung (1990), and then, as with the BEC, each word class was further analysed and placed into semantic categories. Fig. 55 below shows the key words computed from the PMC divided by word class (see Appendices 13 and 14 in Vol. II for a full grammatical and semantic categorisation of these key words). Each entry shows the number of words found in each word class:

Fig.55 Distribution of key words by word class in the PMC (BNC reference)

a) Nouns

The semantic categorisation of the positive key word nouns in the PMC (BNC reference) was very similar to that of the BEC. In all, ten categories were identified, of which nine were the same as in the BEC; people, institutions, activities, things, states & qualities, amounts & measures, places, events and money; and a new category – food. For reference, this is presented below in the diagram:

people food states & qualities

institutions amounts & measures

Positive nouns

activities PMC -> (BNC) places

things money events

Fig. 56 Semantic categorisation of positive key nouns PMC -> BNC

People: This was the second largest category of nouns (66 instances- 21% of the nouns) found and displayed a high level of business-relatedness – 58 people out of 66 instances in the sample. The most ‘key’ were manager, customer, employee, director, personnel and management. The ten non-business people found here were inhabitant, participant, person, homemakers, exhibitor, travelers,[199] westerner, analyst, royalty and technician. Thus, business-related people constituted 87.88% of the sample, comparing favourably to the similar category in the BEC, where business-related people were 86.56% of the sample.

Institutions: This was a small category (14 instances – 4% of the nouns) where the words referred to types of institution, for example, company, organisation, plc, corporation, agency and conglomerate. The corresponding institutions category in the BEC was much larger – 8% of key nouns in the BEC as opposed to 4% in the PMC – and more varied in its contents compared to the examples of institutions found here. This points to a much broader lexical view of business in real-life than is found in the materials.

Activities: The activities (17 instances – 5% of the nouns), like the similar category in the BEC, were either pure business, for example, business, production, investment, downsizing, outplacement, overtime, or words that could fit into the business environment, for example, appraisal, expansion, selection. Also found here were clear non-business activities such as sport and politics.

Things: This was the largest category of nouns (113 instances – 37% of the nouns). It could be divided into tangibles (87 instances – 76% of the category), e.g. product, clothes, insurance, vehicle, computer, and more abstract words (26 instances – 23% of the category) such as idea, culture, opinion. Several points can be raised about the nouns in the things section.

There was a distinct group of business words – product, industry, industry, TQM, GDP and waybill, for example. Likewise, a group of words in the ‘semi-business’ category, i.e. they would not be out of place in business, but were not pure business words, for example, performance, insurance, equipment, objective and software.
There was also a group of completely non-business-related nouns, for example, vehicle, term, moment, name, TV, street, car, boulevard, morning and language.
Significantly, there was a higher concentration of abstract lexis in this category of the PMC when compared to the same category in the BEC. In the BEC things category, there was a ratio of 84% tangible – 16% abstract nouns, and in the PMC, 76% tangible – 23% abstract nouns.

States & qualities: This PMC category was relatively smaller than the corresponding BEC category – (PMC: 16 instances – 5% of key nouns sample / BEC: 43 instances – 9% of sample) – and showed a mixture of both positive and negative states: success, pleasure, growth, as opposed to inflation, bankruptcy, redundant and inconvenience. Two key differences can be discerned between the nouns in this category of the PMC and those in the BEC. Firstly, as noted above, the difference category size. Secondly, the BEC displayed overwhelmingly positive/neutral states and qualities – (84% of the sample) – with only 6 nouns being negative in connotation (16%). In the PMC, negative states accounted for 25% of the sample. This may be indicative of an over-emphasis in the materials on the problems faced in business that, whilst certainly present in real business life, would not occur in such a high concentration. An example is shown below using the word inconvenience, where it is clearly being taught for both written and spoken situations as part of the function of apologising:

Three categories that were found in both corpora (PMC and BEC) – measures & amounts, places and events, were all quite similar in size. However, the biggest of these three categories in the BEC was events, whilst in the PMC it was places.

Measures & Amounts: It was noted in the BEC that the only key amounts were at the high end – million, billion, trillion – and the most key noun in the measures category in the BEC was billion. In the PMC, the most key word in this category was million and, although billion was also key in the PMC, trillion was not. This suggests that examples of money-related matters tend to be kept to lower, more everyday amounts in published materials, whereas in real business life, much higher numbers are more common. This is confirmed by that fact that the word hundreds was a positive key word in the PMC, and hundred was, in fact, a negative key word in the BEC.

In the places section, whilst many words were purely business-related – department, office, factory – there were also several words where the entertaining and travel aspects of business came to the fore – hotel, restaurant, airport – and some words with no real business connection, for instance, crèche, metro and country. The events category is notable for the number of negative events – theft, recession, breakdown, downturn – which although few, actually make up 28% of the category. This was not found in the BEC, where only one[200] clearly negative event was found – bankruptcy (4.34% of the category).

Thus, analysis of these three categories showed the following.

In the PMC there is an emphasis on naming places – 6.57% of nouns in the PMC are related to place names, compared to only 3.2% in the BEC.
In the PMC there is slightly less emphasis on events – 4% of nouns in the PMC related to events compared to 5% in the BEC.
Negative events nouns are more featured in the PMC than in the BEC – 28% to 4%.

The remaining categories, food and money, showed a preoccupation in the materials with tangible situations.

Food: The inclusion of food-related lexis, for example, pasta, sauce, dessert and chocolate, indicated once again the focus on entertaining in the materials. Examples of the usage of the word meal can be seen here:

Money: There was slightly less focus in the PMC on money/finance than was found in the key nouns of the BEC. Reference to money matters in the PMC amounted to only 5% of the key nouns, whereas in the BEC it constituted 9% This was a surprising result, as money is such a central part of business, and an aspect of the business world that materials writers must all have personal experience of.

In addition to the categories considered so far, there was also one semantic category of key nouns that was present in the BEC, but missing from the PMC – technology & computers. This was a relatively small category in the BEC – just 4% of key nouns – but its absence from the PMC can be seen as significant. The absence is, at least, partially explained by the fact that the most recent book in the PMC was published in 1996, meaning that the impact of computers had not had time to reach the materials. However, it is another example of the lexical limitation of the PMC, where a key aspect of business life is not reflected in the materials.[201]

In summary, several points can be raised about the nouns in the PMC.

The nouns form similar semantic groups in the PMC compared to those found in the BEC, but there is greater focus in the PMC on the lexis of entertaining and travel. This, however, is not surprising in itself, as entertaining/travel situations were not part of the BEC.
The nouns in the BEC were more evenly spread throughout the categories. In the PMC two main categories – people and things – accounted for the majority of the nouns (60%). Consequently, there was less lexical diversity in the PMC than the BEC, for example, shown by the institutions category (4% of the key nouns sample in the PMC and 8% in the BEC).
The key nouns in the PMC were mostly tangible, and the amount of states and qualities referred to by the nouns was slightly less in the PMC (5%) than in the BEC (9%). However, more abstract lexis was found in the things section of the PMC compared to the corresponding category in the BEC.
More negative lexis was found in these nouns than in the corresponding sections in the BEC. This is exemplified by the events category where 28% of PMC events were negative (e.g. theft, recession, breakdown), matched to only 4% of BEC events.
Measures & amounts found to be key here referred to lower numbers (e.g. hundreds, million) than those found to be key in the BEC. This suggests that the PMC has brought numbers more down to the ‘everyday’ level, whereas the BEC focuses more on high-level numbers related, perhaps, to big business.
The PMC focused less on key words associated with money/finance – only 5% of PMC key nouns compared with 9% of BEC key nouns.

b) Verbs

TABLE LXV: SEMANTIC CATEGORISATION OF PMC KEY WORD VERBS

Negative	Neutral	Work/ Business	*Possibility/* *Necessity*	Inter personal	Personal	Running a business
complicates	arrive	advertise	could	please	prefer	send
plummet	prepare	enclose	would	thank	enjoy	arrange
	let	sell	needn’t	agree		receive
Positive	make	manage	should	meet		accept
	underlie	confirm	mustn’t	discuss		send
excite	accompanies	deliver		speak		develop
expand	wiped	invest	Technology	introduce		specialize
appreciate	settle	install		inform		locate
improve		employ	automate	advise		operate
solve		repay		recommend		unload
		hire		motivate		redesign
		despatch		expect		reorganize
		diversify		summarise		recycle
		promote		interrupt
				explain
				suggest
				mis- understand
				ask
				instruct

If the key nouns found in the PMC could be considered similar to those in the BEC, the same could not be said of the verbs to the same extent. In all, nine semantic categories were found for the key verbs of the PMC, eight of which were also found in the BEC: negative, neutral, work/business, technology, running a business, positive, interpersonal and personal but also a new category – possibility/necessity. The verbs are shown in full above with the new category shown in italics.

There are, indeed, similarities between the PMC key verbs here and those in the BEC. As was seen above, eight of the semantic categories can be related to similar categories in the BEC. However, there are several important differences between the verbs found here and those in the key verb category in the BEC.

In the BEC, the verbs in the running a business category constituted by far the biggest category (38 instances – 40% of the verbs). When these verbs were added to the second largest category of BEC verbs – work/business – the total of business-related key verbs in the BEC accounted for 65.59% of the sample. In contrast, in the PMC, business-related verbs amounted to a little under 40%. Thus, verbs that are actually of central importance to learners, pure business verbs, e.g. negotiate, shelve, merge and sub-business verbs, i.e. verbs to describe the surrounding activities/actions of business, e.g. comply, propose, expand, ensure – are lacking in both number and significance in the materials.

In the PMC, verbs concerned with interpersonal communication constitute the single largest group of key verbs (19 instances – 27% of the sample), e.g. please, thank, agree, meet. This contrasts with BEC key verbs, where the corresponding interpersonal category amounted to only 10% of the sample. Thus, the private verbs of Biber (1988) – relating to personal and emotive states – are stressed here, where they were not stressed in the BEC positive key verbs. This theme of materials’ over-representation of personal lexis will be returned to later in this section.

The verbs also displayed a preoccupation with possibility and necessity with inclusion of modals could, would, should and negatives needn’t and mustn’t. The fact that these verbs were key results at least partly from the following:

a focus on requests (here using the example of the word could);

on polite phrases, also including requests found in the materials;

and business possibilities, for example, in negotiations;

Thus, business lexis found in the BEC used these types of verbs significantly less and the PMC used the same types of verbs specifically more. The key verbs in the PMC, therefore, present a different reality from that found in the BEC.

In summary of the verbs section, the following points can be made.

The verbs in the PMC were more concerned with the personal and interpersonal. It will be also remembered that it was just these categories (personal, interpersonal) that were so central to the negative key words in the BEC. This means that those aspects of lexis deliberately used less in the business world as presented by the BEC are deliberately used more in the PMC.
There is a strong focus on the possible and polite.
Whilst business-related verbs in total represent the biggest group of verbs – i.e. if the work/business and running a business categories are added together – they are still, proportionally, 26% less than the same categories in the BEC key verbs.

c) Adjectives

The semantic categories of adjectives employed in the BEC were again used in the PMC: size/speed, places, positive, negative, neutral, work/business, money, technology and time. There were similarities between the two sets of adjectives, for example, the proportion of key adjectives in each category was similar (9% of all key words in the PMC to 11% in the BEC). However, there were also differences. There were proportionally more negative adjectives in the PMC (7% of the sample) than the BEC (2% of the sample). This negative PMC category included two quite personal adjectives, tiring and depressing,[202] which are in contrast to the non-emotive language of the BEC. The two other negative adjectives, afraid and sorry, appear for very specific reasons. Afraid is found in the PMC almost entirely in the combination I’m afraid as a warning discourse marker of bad news coming (349 instances out of 358 – 97% of all occurrences in 30 books):

Likewise, sorry appears largely as part of the materials’ preoccupation with apologising and politeness:

The adjectives in the PMC, then, fall into the same semantic categories as the BEC, but are much more limited in scope. The highest frequencies of adjectives were taken up by the repetition of certain set phrases, for example, I’m sorry to trouble you, I’m afraid, that do occur in real life, but are here over-stressed in relation to reality – at least the version of business reality presented by the BEC.

d) Noun/verbs

As with the adjectives, the semantic groups for noun/verbs found in the BEC could be applied to the PMC: places & people, communication, positive, negative, neutral, work/business, money, technology, running a business and time.

In the PMC, the work/business noun/verbs were the largest group (35 instances – 23% of the category), whereas in the BEC it was neutral noun/verbs that represented the largest group (53 instances – 26% of the category).
The running a business group of noun/verbs was proportionally slightly smaller in the PMC than in the BEC (16% in the PMC to 17% in the BEC). In the PMC there were more negative noun/verbs than positive (6 negative – 5 positive), in the BEC this was the opposite way around (4 positive – 3 negative). Thus, again there were proportionally more negative attributes noted by the PMC (6 instances) as opposed to only 3 instances in the BEC, which was double the size.
Again in evidence was a large number of noun/verbs related to money and finance, showing that they have both a nominal and verbal function in Business English.
In all, this group was relatively similar to that found in the BEC and showed a high degree of business-related lexis. In fact overall, business-related lexis accounted for 39% of the group (as opposed to 35% in the BEC), and there was only slightly less sub-business lexis than that found in the BEC.

The noun/verb PMC-BEC differences are summarised in the table below:

TABLE LXVI: NOUN/VERB PMC-BEC DIFFERENCES

PMC nouns/verbs	BEC nouns/verbs
Largest group: work/business 35 – 23%	Largest group: neutral 52 – 26%
Running business category: smaller – 16%	Running business category: larger 17%
More negative than positive n/v: 6 negative words	More positive than negative n/v: 3 negative words
Overall more business-related lexis: 39%	Overall less business-related lexis: 35%

9.4.1.2 Discussion: results of Analysis 1

The main title of this section asked how key words computed from the PMC lexically demarcated the world of business. The world of business created by the PMC key words can be seen in the diagram below:

people personal positives

money PMC negatives

Business

things English interpersonal

running a business politeness

institutions measurements & amounts travel/entertaining

Fig. 57 The lexical world of Business English as found in the PMC (BNC reference corpus)

Semantically, the groups formed from the key words in the PMC in Analysis 1 were focused on business or business-related matters, and there was a stress on travel and entertaining, interpersonal communication and a relatively even distribution between positive and negative lexis. There was also a stress on the possible and the polite. When compared to the key word categories found for the BEC the similarities and differences can be summarised as follows.

Semantic groups: Similar semantic groups were identified into which the key lexis of the PMC could be fitted.
Personal and interpersonal lexis: The key lexis in the PMC was more concerned with personal and interpersonal lexis, especially in the verbs category – thank, agree, discuss, prefer, enjoy – than key lexis found in the BEC.
Positive and negative lexis: The balance between positive and negative lexis in the PMC is in contrast to the BEC, where more positive language was found.
Tangible vs abstract: In one large noun category – things – there was slightly less emphasis on tangibles in the PMC than in the BEC – though both sets of key words are very tangible-oriented. Conversely, slightly fewer states & qualities were expressed in the PMC than in the BEC.
Verbs vs noun/verbs: The PMC displayed fewer verbs related to work/business and running a business (39% of the PMC sample as opposed to 65.59% of the same category in the BEC). However, the PMC used proportionally more noun/verbs for the same purpose. Whilst there were slightly more pure business noun/verbs in the BEC (BEC 17% – PMC 16%), overall business-related noun-verbs were more frequent in the PMC than in the BEC (PMC 39% – BEC 35%). Thus, proportionally, the PMC tended to express business concepts more through use of noun/verbs and less through use of verbs. In contrast, the BEC used noun/verbs slightly less to express business-related lexis, but used pure verbs more.

The fact that published materials express business-related ideas more through noun/verbs, whilst real business favours pure verbs, is interesting. It may represent a narrower base of business activities in the materials, or it may reflect the predilection of materials writers on naming aspects of business, as opposed to the language needed for doing business. This is reinforced by Lewis’s (1993:39) observation that ‘Language teachers, usually accidentally, see vocabulary largely in terms of nouns’. Lewis admitted (personal communication 2000) that his observation was only based on his personal experience and not empirical analysis, but the comparative data of the PMC and BEC confirms his observation. This would suggest that there needs to be a greater focus on verbs in Business English teaching materials than is found at present.

Throughout this last section, reference has been made back to the key words of the BEC and how the key words of the PMC compared to them when both sets of key words were computed using the same reference corpus (the BNC). The next stage of the study continues with Analysis 2, by studying the key words found in the PMC computed by statistical reference to the BEC, rather than the BNC. After this analysis, conclusions can be reached as to the main lexical differences between the PMC and the BEC.

9.4.1.3 Analysis 2:

Key word analysis of the PMC (with the BEC as the reference corpus)

The analysis carried out above examined the key words of Business English found in the PMC. It thus presented the world of business lexis as seen through the eyes of materials writers. The second key word analysis carried out, took as the starting point two sets of Business English: real-life Business English (the BEC) and the Business English found in the PMC. By statistically comparing the two corpora, it was possible to see which aspects of business lexis were stressed in the PMC, i.e. what materials writers consider important language to include in their books. This part of the study is similar to that carried out by Ljung (1990), when he compared his corpus of EFL written teaching materials to the COBUILD corpus. Ljung used frequency as the basis for analysis, that is, he took the top 1,000 words of the two corpora and compared frequency of use in his corpus to that found in COBUILD. He established a difference coefficient (based on the work of Yule 1944) to show if a word had been over- or under-represented in the materials corpus.

Whilst Ljung’s work was valuable, and helped form the structure of this thesis, it was felt at the outset that a focus on key words rather than frequency would produce more reliable, and hopefully, interesting results. The results of the comparison of the PMC to the BEC show, therefore, the lexis that is over-represented in the PMC – the lexis that occurs to an unusual level of frequency as compared to the BEC.[203] Thus the main lexical interest and focus of the PMC could be determined.

It was stated in Hypothesis 2 that the two sets of lexis formed from the two corpora would be significantly different. This was indeed found to be the case, and the rest of this section exemplifies and discusses the differences found.

As with the previous analyses above, the PMC corpus was lemmatised using the modified lemma list of Someya (1998) and was statistically compared to the lemmatised BEC using the log likelihood statistic (Dunning 1993), with a p value of p = 0.000001. A key word list was generated and manually edited to sift out names of people, products, places, currencies, and companies. The full key word list can be found on the CD ROM in the back page of this thesis. The top 100 key words are printed below:

TABLE LXVII: THE 100 MOST ‘KEY’ KEY WORDS OF THE PMC (BEC REFERENCE CORPUS)

N	WORD	PMC FREQ.	PMC.LST %	BEC FREQ.	BEC.LST %	KEYNESS Log L.
1	YOU	10 587	1.78	10 133	0.99	1 800.6
2	I’M	1 759	0.30	930	0.09	910.3
3	YOUR	2 949	0.50	2 409	0.24	744.2
4	I’D	634	0.11	129	0.01	695.8
5	AFRAID	358	0.06	38		502.1
6	ME	1 399	0.24	978	0.10	479.4
7	OUR	2 539	0.43	2 342	0.23	474.7
8	MY	1 240	0.21	861	0.08	429.9
9	THANK	877	0.15	502	0.05	409.1
10	PLEASE	1 079	0.18	715	0.07	404.7
11	SORRY	447	0.08	166	0.02	331.9
12	LIKE	1 666	0.28	1 555	0.15	301.1
13	I’LL	721	0.12	461	0.05	286.2
14	WE	5 837	0.98	7 492	0.73	284.0
15	FLIGHT	225	0.04	38		268.6
16	COULD	1 317	0.22	1 176	0.11	268.0
17	ABOUT	2 096	0.35	2 222	0.22	252.3
18	GOOD	1 456	0.25	1 386	0.14	248.7
19	SEE	1 398	0.24	1 360	0.13	223.9
20	HOW	1 385	0.23	1 351	0.13	220.0
21	CAN	2 545	0.43	2 947	0.29	213.9
22	TELEX	167	0.03	25		209.1
23	YOURS	302	0.05	134	0.01	190.0
24	LETTER	356	0.06	185	0.02	187.9
25	LTD	332	0.06	163	0.02	187.3
26	JOB	755	0.13	627	0.06	183.1
27	MEET	456	0.08	294	0.03	178.6
28	MRS	161	0.03	33		176.0
29	MORNING	328	0.06	174	0.02	168.8
30	GOODBYE	103	0.02	6		165.5
31	MR	1 041	0.18	1 025	0.10	160.5
32	MANAGER	811	0.14	742	0.07	154.7
33	LET’S	274	0.05	135	0.01	154.0
34	PERSONNEL	219	0.04	90		148.5
35	LUNCH	167	0.03	52		142.3
36	TELL	527	0.09	419	0.04	140.7
37	DEPARTMENT	456	0.08	339	0.03	139.4
38	SINCERELY	204	0.03	84		138.1
39	HELLO	303	0.05	177	0.02	137.4
40	OKAY	89	0.02	7		134.7
41	FINE	315	0.05	193	0.02	133.4
42	ADVERTISE	274	0.05	154	0.02	130.9
43	SURE	501	0.08	403	0.04	130.4
44	NAME	510	0.09	414	0.04	130.1
45	SIR	191	0.03	81		125.7
46	US	1 186	0.20	1 300	0.13	125.6
47	DRINK	180	0.03	72		125.2
48	EMPLOYEE	411	0.07	307	0.03	124.5
49	AIRPORT	122	0.02	29		123.4
50	MACHINE	293	0.05	182	0.02	121.5
51	INTERVIEW	197	0.03	92		117.5
52	OH	544	0.09	475	0.05	117.1
53	CHOCOLATE	69	0.01	3		116.1
54	NICE	204	0.03	100		115.3
55	HEAR	271	0.05	167	0.02	113.7
56	TOO	520	0.09	452	0.04	113.2
57	PRODUCTION	362	0.06	267	0.03	112.3
58	COURSE	591	0.10	546	0.05	109.8
59	CONSIGNMENT	91	0.02	15		109.7
60	AH	134	0.02	45		107.9
61	POINT	692	0.12	679	0.07	107.8
62	ARRIVE	208	0.04	112	0.01	105.0
63	LEAVE	436	0.07	369	0.04	101.2
64	FIRST	919	0.15	1 011	0.10	95.9
65	ENCLOSE	186	0.03	99		95.3
66	IDEA	427	0.07	369	0.04	94.2
67	VERY	1 362	0.23	1 642	0.16	94.0
68	QUESTION	454	0.08	404	0.04	93.1
69	LET	373	0.06	306	0.03	93.0
70	LOOK	1 045	0.18	1 197	0.12	92.0
71	ASK	539	0.09	515	0.05	91.0
72	MISS	146	0.02	66		90.1
73	HOUR	350	0.06	285	0.03	88.7
74	PREMISE	44	0			88.2
75	MMM	44	0			88.2
76	SPEAK	351	0.06	288	0.03	87.4
77	TAKE	1 269	0.21	1 543	0.15	83.8
78	MEETING	697	0.12	739	0.07	83.7
79	FACTORY	199	0.03	123	0.01	83.1
80	ENJOY	163	0.03	88		82.0
81	HERE	719	0.12	773	0.08	82.0
82	PRESENTATION	168	0.03	93		81.9
83	WORKER	176	0.03	102		80.7
84	FIGURE	326	0.05	269	0.03	80.2
85	EXCUSE	91	0.02	27		80.2
86	AFTERNOON	146	0.02	74		79.4
87	MANPOWER	62	0.01	9		78.5
88	DEAR	360	0.06	316	0.03	76.4
89	ROAD	231	0.04	165	0.02	76.1
90	AGREE	482	0.08	472	0.05	75.6
91	INTEREST	716	0.12	789	0.08	74.2
92	CAN’T	408	0.07	381	0.04	73.5
93	MEAL	75	0.01	19		73.1
94	STAFF	437	0.07	419	0.04	73.0
95	DISCOUNT	225	0.04	163	0.02	72.2
96	ADVERTISEMENT	65	0.01	13		71.9
97	FAITHFULLY	79	0.01	23		70.5
98	SEAT	91	0.02	33		68.9
99	SOON	249	0.04	199	0.02	65.7
100	CANDIDATE	132	0.02	72		65.6

It can be noticed immediately that this list is significantly different from both the BEC key word list seen in Section 9.3.1.2 and the PMC (BNC reference) key word list noted earlier in Section 9.4.1.1. If it can be assumed that both the PMC and the BEC are rich in business lexis, the frequency of this lexis should cancel itself out. That is, if the business lexis is frequent in both corpora it will not show up as statistically key. What is left, then, is the lexis that is significantly more frequent in the PMC than in the BEC. The business lexis found to be key here was, for example, ltd, job, manager, personnel, department, advertise, employee, production, consignment, factory, worker, manpower, staff and advertisement. Thus, these are the business words that appear significantly more in the PMC than in the BEC, showing the materials writers’ favoured business lexis. More apparent than this, though, is a preponderance of the lexis of politeness, travel, correspondence and interpersonal communication. In order to further analyse this lexis, the key words were categorised by word class using the system of Ljung (1990), shown below:

Fig. 58 Distribution of key words by word class in the PMC (BEC reference)

Then, as with the BEC, each word class was further analysed and placed into semantic categories. These will now be looked at briefly in turn (see Appendices 15 and 16 in Vol. II for full grammatical and semantic categorisation of the PMC – BEC reference – key words).

a) Nouns

The nouns in this part of the analysis could be divided into semantic groups similar to those of the BEC, but with three new with new categories – travel, time and food – shown below in italics:

people institutions time

things Key Nouns activities & events

travel states & qualities food places

Fig. 59 Semantic categorisation of PMC key word nouns (BEC reference)

The nouns, and indeed the other lexis in the PMC, exhibited a consistent focus on certain limited aspects of business life. Whilst six of the nominal semantic categories found in the PMC were the same as in the BEC, they were much reduced in size, and an over-representation of time, travel and food was found.

People: This was the second largest group of key nouns in the PMC (35 words – 24%). It showed both a large number of business-related people – manager, personnel, employee, worker, director (22 words- 62% of the category sample) – but also a stress on many non-business-related people (9 words – 25% of the category sample), e.g. inhabitant, woman, population, homemakers, wife, grocers, son, papa and doctor. Also present were a number of titles related to people – Mrs, Mr, sir, ma’am and madam, with Mrs and Mr being the most key references to people in the sample. The materials, then, see this area of business in terms of certain key jobs – for example, manager, director, supervisor and importer – and in terms of the correct form of addressing the people in both spoken and written situations:

It is interesting to note that the materials make a distinct effort to mention women, with Mrs, ma’am and madam all being key words in the PMC. Even where the business situation is predominantly male, women are brought into the conversation. The result of this female focus, however, does not eradicate sexism in the PMC, but often continues to place women in a non-business environment, as can be seen in the examples below:

What is significant about the nouns in general is the fact that aspects of the negative key nouns in the BEC are positive key words in the PMC. The analysis of the BEC showed that Business English is lexically demarcated from lexis concerned with family and society, i.e. family and society-related lexis occurs far less in Business English than in everyday English. The PMC, in contrast, over-represents this language, e.g. wife, son, doctor, papa. It seems to be a result of materials trying to contextualise situations and placing emphasis on social situations.

Only three institutions were key: EC, company and EEC, and whilst the activities & events category was also small (15 nouns – 10% of key nouns) it showed the focus of Business English materials in terms of what they consider happens often in business well, for example, production, meeting, presentation, franchising, promotion, recession, shipment and recruitment. Also present were non-business-related nouns – theft, accident and birth. These negative events – theft, accident – continue the theme noted in the earlier analysis, whereby negative aspects were much more heavily present in the PMC than in the BEC. It may be that an emphasis on dramatic situations in the materials is a result of attempts to liven up the materials for students:

Things: This was the largest group of key nouns in the PMC (38 – 26% of key nouns) and displayed a high degree of tangibles: letter, job, consignment, advertisement. Only two possible intangible nouns could be identified – idea and tactic – meaning that in this category 95% of the nouns were tangible. Business and non-business-related items were both present. The business items again show the business world as presented by writers: letter, job, advertisement, flexitime, cheque, fringe (benefits), employment, product, quota and memo. There is also a number of non-business lexis here – idea, diamond, documentary, furniture, sweater, scarf, perfume and cliff – though arguably some of these are products in their own right. Also included in the non-business lexis was the word problem. Thus the preoccupation with problems continues:

The two remaining categories that were the same in the BEC, states & qualities and places, were both small.

States & qualities: This category consisted of only 3 nouns – pleasure, probability and maternity – amounting to only 2% of the key nouns. This contrasts with 43 states & qualities (9.7% of the sample) found in the BEC key word nouns. This distinct lack of lexis relating to states and qualities, and the over representation of tangibles noted earlier, shows features discussed by Ljung (1990). Ljung, in his comparison of EFL materials and real-life English – from COBUILD – had noted an under-representation in the EFL materials of the abstract, and an over-emphasis on the concrete: ‘the proportions between the abstract and concrete, between complicated and simple, have been skewed in the direction of the concrete and simple’ (Ljung 1990:17).

Places: The places category of the PMC focused on typical places of business, for example, department, factory, office and warehouse, and also on aspects of food/entertaining, for example, canteen, theatre.

Two of the three new nominal categories in the PMC continued this food/entertaining theme to distinguish central aspects of Business English materials – time, travel and food. This resulted from the fact that virtually all the books contained reference to business entertaining and travel. As has already been stated, no entertaining or travel situations were included in the BEC, so a direct comparison is not possible, but the fact that these semantic categories appeared – even when using the BNC as a reference point food formed a distinct semantic group in the PMC nouns – helps us to gain an overall view of business as seen through the eyes of materials writers.

In summary of the noun section, a picture of the lexical business world found in materials started to emerge. In the PMC a typical business person is a manager, employee or worker. They work in a department, factory or office and their key activities are concerned with production, courses, meetings and presentations. Key events include promotion, and time, food and travel are central to their life. This is summarised in the diagram below:

Fig.60 The business world found in the key nouns of Business English teaching materials

This is, of course, an over-simplification, but it does give an overview of the semantic categories found in the PMC and some of the lexis contained within them.

b) Verbs

As these were quite few in number they can be shown in full:

TABLE LXVIII: PMC POSITIVE KEY VERBS (BEC REFERENCE)

Neutral	Inter- personal	Work/ Business	Negative & Modals	Personal	Technology	Running a business
do	thank	advertise	could	see	automate	enclose
settle	please		can’t	arrive		arrange
accompanies	meet		would	let		specialize
eat	tell		mustn’t	take
wiped	hear		needn’t	enjoy
	ask		don’t	dislike
	speak		won’t	prefer
	agree			accept
	interrupt			appreciate
	apologize
	introduce
	recap

The positive key verbs of the PMC shown above are similar to the negative key verb categorisations of the BEC, shown here again for comparison:

TABLE LXIX: BEC NEGATIVE KEY VERBS

Negatives & Modals	Neutral	Personal	Interpersonal
didn’t	get	know	say
can’t	gonna	see	tell
couldn’t	do	eat	hear
haven’t	come	remember	lie
wasn’t	put	die	marry
could	read	pray	elect
won’t	wear	sit	listen
weren’t	hang	think	ask
wouldn’t	burn	wanna	pretend
isn’t	fetch	feel
hasn’t	have	suppose
	seem	born
		tire
		forget
		condemn
		reckon
		observe

It can be seen that three categories are the same in the two sets of data – personal, interpersonal and negatives & modals. Thus, the language being stressed in the materials is focused around personal and interpersonal verbs, for example, know, see, eat, remember, arrive, let, take, and enjoy. There is also a focus on possibility and necessity, e.g. could, can, can’t and would. This may be explained by the fact that the writers have attempted to personalise the materials – that is, a link is created between the student and the business situations by showing on-going dialogues/situations between people. An example is shown below with the word remember:

This over-representation of personal and interpersonal lexis contrasts with a low number of PMC verbs related to business. Only 4 verbs were found – advertise, enclose, arrange and specialize – which is in marked contrast to the key verbs of the BEC where business-related verbs accounted for 66% of the sample. In sum, the studies carried out here on verbal usage in the PMC find it to be distinctly lacking in sub-business vocabulary – which common sense dictates should be the verbs to be stressed, if not over-stressed. Instead, a high number of personal and interpersonal verbs, and those often used in polite phrases, e.g. can, could, would, were the main focus of attention.

c) Adjectives

The notable feature of the BEC positive key word adjectives was the number of words with a positive semantic connotation, for example, new, best, successful, available and relevant. Here, in the PMC, as seen below, the largest group of adjectives is negative:

TABLE LXX: PMC POSITIVE KEY ADJECTIVES (BEC REFERENCE)

Places	Positive	Negative	Neutral	Work/ Business
abroad	good	afraid	sure	LTD
	nice	sorry	main	irrevocable
	new	expensive	laden
	grateful	tiring	true
	glad	late
	convenient	depressing
		bad

Moreover, the negative adjectives and indeed the positive, also mostly fall into the emotive bracket – afraid, sorry, tiring, depressing and bad; good, nice and glad – which were absent from the BEC. As was discussed with the PMC key adjectives previously, these words are key as a result of constant repetition in letters and social situations – afraid, sorry, grateful, convenient, that exist in real life – but are over-stressed in the materials:

d) Noun/verb

The overall similarity in semantic categories found in both the PMC and BEC was also present in the noun/verb category. Ten categories were identified in the PMC:

TABLE LXXI: PMC POSITIVE KEY NOUN/VERBS (BEC REFERENCE)

Places & People	Food & Drink	Travel	Negative	Neutral	Personal/ Inter-personal	Work/ Business
staff	lunch	ski	regret	can	like	telex
police	drink	trip	steal	name	leave	interview
park	bottle	travel		point	question	interest
referee	milk	sail		seat	look	discount
clerk	cream			sparkle	miss	brand
	fry			hand	excuse	invoice
Time	dining	Running a business		plant	hope	ENC
	sugar			box	answer	career
time	breakfast	figure		dress	stay	credit
		call		break	mind	price
Technology		campaign		tile	reply	overwork
		draft		divide	spell	import
machine		visit		present	feel	DEL
				sketch	read

The business-related noun/verbs found in the PMC again offered the layperson’s view of business, for example, telex (somewhat dated now), interview, interest, discount, brand, invoice, career, credit and price. The fact that the published materials related to business through noun/verbs rather than pure verbs is found again here. Only four business verbs were found to be key in Analysis 2, whereas 18 noun/verbs relating to business were identified. In addition to these business-related words was the lexis of food & drink: lunch, drink, bottle, milk; the lexis of travel; trip, travel, sail, and personal and interpersonal lexis; question, look, miss, excuse. These findings correlate with all the others found in this section of the analysis with a focus on food, drink, travel and a limited range of business activities. These key words also differ markedly to those noun/verbs seen for the PMC in the last section, when the BNC was used as reference corpus and the focus noted then for this category on business-related lexis is here largely superseded by neutral, personal and interpersonal noun/verbs.

9.4.1.4 Discussion: results of Analysis 1 and Analysis 2

It is now possible to give an overall answer to the question posed at the beginning of the analysis of the PMC – how do the PMC key words define the lexical world of business and how does this definition compare to that shown in the BEC?

Analysis 1: Analysis of the PMC – with the BNC as the reference corpus – showed that in many ways the materials demarcate the lexis of Business English in a similar manner to the BEC, as semantic categories common to both were found. The most ‘key’ words in both were also very similar, for example, business, company, market and sales. Differences, however, were found in the usage of personal and interpersonal lexis, the balance between positive and negative lexis, the amount of tangible and abstract lexis and the purpose to which pure verbs and noun/verbs was put. These were all summarised in full at the end of Analysis 1.

Analysis 2: Analysis of the PMC – with the BEC as the reference corpus – showed which lexical aspects of the published materials are over-represented when compared to the BEC. The differences found in Analysis 2 are noted below.

The lexis of family and society is over-represented in the PMC, e.g. wife, son, doctor.
There is a focus on a limited number of job-related terminology, e.g. manager, personnel, employee.
The PMC gives more focus on women than the BEC, and several lexical items related to women are key, e.g. Mrs, ma’am and madam.
Tangible lexis is more heavily represented in the PMC in comparison to the BEC. For example, in the PMC, the key noun category of things consists of 95% tangible nouns, compared to 84% tangible nouns in the same category in the BEC.
The lexis of food, entertaining and travel is far more prominent in the PMC than in the BEC. This is shown by the fact that none of these semantic categories were formed in the BEC.
Those verbs over-represented in the PMC are concerned with personal and interpersonal matters – thank, see, arrive, enjoy, dislike – and only four business-related or sub-business verbs are over-represented – advertise, enclose, arrange, specialize.
Negativity is over-represented in the PMC. This is shown in both verbs – didn’t, can’t, couldn’t – and in adjectives – afraid, sorry, expensive, tiring.
As with Analysis 1, noun/verbs are used to refer to business matters more than pure verbs (4 business-related verbs to 18 noun/verbs).

Both analyses: It was found, therefore, that trends in the lexis of the PMC that were noted in Analysis 1 became even more pronounced in Analysis 2. Taking the two analyses together, the differences found between the PMC and the BEC can be summarised as follows, showing the differences between the view of business of the materials writers and real-life:

TABLE LXXII: DIFFERENCES BETWEEN KEY WORDS FOUND IN THE PMC AND THE BEC

PMC key features	BEC key features
1. Negative & positive lexis	1. Stress on positive lexis
2. Focus on personal & interpersonal lexis	2. Focus on non-personal lexis
3. Limited states & qualities	3. Wider range of states & qualities
4. Stress on polite language	4. No stress on polite language
5. Less pure business lexis	5. High level of pure business lexis
6. Less sub-business lexis	6. High level of sub-business lexis
7. Emotive adjectives	7. Non-emotive adjectives
8. Strong focus on tangibles	8. Focus on tangibles
9. Problems	9. Solutions
10. Business concepts expressed more through noun/verbs	10. Business concepts expressed more though verbs

9.4.2 Further analysis – PMC negative key words

In the analysis of the BEC, both positive and negative key words were used to demarcate the lexis of business.[204] Analysis of the PMC has so far only used the positive key words. However, whilst no systematic analysis of the negative key words of the PMC was carried out, a brief overview can add interesting dimensions to the study.

The negative key words of the PMC (BEC reference corpus) were printed out and analysed (the full list can be found in Appendix 18 in Vol II). The most negative key words, perhaps surprisingly, contained a high number of pure business words, for example, contractor, seller, tax, business, service, audit, budget, investor, distributor, stock, financial, merger and consortium. Also found were a large number of sub-business words, for example, review, report, support, provide, focus, system, design and strategic.
Additionally, there was lexis referring to states and qualities, for example, growth, leadership, individual, consolidation, independent, professional and completion.

There is seemingly, therefore, a reversal – what was positive in the BEC is negative in the PMC and vice versa. Many of these words were actually key in the PMC using the BNC as reference, but here they are placed as negative key words. Thus, although they were key in the PMC-> BNC comparison, they were not key enough in the PMC->BEC analysis. Whilst the presence of some of the words in the negative column – contractor, seller, for example – can be explained by their total frequency being high in the BEC due to some skewing of the corpus, many of the words cannot, for example, business, service, audit and financial. This may suggest that the materials under-represent lexis that is central to business life, but further research would be needed to confirm this.

This would be significant, and may show that whilst the PMC as would be expected contains much more business-related lexis than the BNC, the business lexis it has is both less diverse and lower in relative frequency than that of the BEC. The negative key words found here may confirm, therefore, the analyses of positive key words of the PMC, where less business-related lexis was also found in relation to the BEC.

Discussion so far has viewed the lexis of the PMC to a large extent at a macro-level. The next section of analysis looks at the lexis at the level of collocation and semantic prosody.

9.4.3 Semantic prosody in the PMC

In order to study the phenomenon of semantic prosody in the PMC, five words that were analysed in the BEC were chosen for further analysis in the PMC – manager, customer, product, market and business. These words were all chosen as being the most ‘key’ key words in the PMC when compared to the BNC. In general, it was found that the basic semantic categories the words collocated with were very similar to the those they collocated with in the BEC, but that there were several differences. Each word will now be looked at in turn.

Manager: The first point to be made is the relative frequency of manager in both corpora. In the BEC it occurred 742 times and in the PMC, 811.[205] As the PMC is only approximately half the size of the BEC, this over-emphasis on the word needs to be stressed from the outset.[206] The semantic prosodies of manager can be seen in Table LXXIII on the next page.

As can be seen in the table, in addition to the categories found in the BEC, the examples from the PMC show another category – name of person + manager. This has arisen as a result of examples of business correspondence in the published materials, where business letters are concluded by the person’s name and the term manager. In both corpora, the largest semantic prosodic group was that of titles. What is not obvious from the table above is that in the PMC, a small number of job titles accounted for a large part of the occurrences. Examples were sales manager (83 examples – 16.02% of the total sample), personnel manager (29 examples) and marketing manager (21 examples), whereas the titles in the BEC were more varied. Despite these differences, it can be said that manager is quite accurately represented in terms of semantic prosody in the PMC, with only a slight lack of diversity of collocating lexis.

TABLE LXXIII: SEMANTIC PROSODIES OF ‘MANAGER’ IN THE BEC AND PMC

Right only:

BEC PMC

semantic prosody	frequency/ 415 & %	example	semantic prosody	frequency/ 518 & %	example
titles	293 – 70.6%	bank manager construction manager R&D manager	titles	365 -70.46%	advertising manager bank manger benefits manager
manager’s qualities (positive)	11 – 2.65%	forthright, excellent manager a good manager	name of person + manager	44 – 8.49%	D. Adair Manager Mrs Helen Tomlinson. Manager
manager’s origins/ nationality	7 – 1.68%	a British manager an ex-pat manager	hierarchy	11 – 2.12%	assistant manager senior manager
company hierarchy	13 – 3.13%	senior manager co-lead manager	positive	14 – 2.7%	a good manager a great manager
			manager’s origins /nationality	7 – 1.35%	Spanish manager Japanese manager

Customer: This word associated with six semantic groups in the BEC, whereas in the PMC only three were apparent. However, the largest categories were found to be present in both, and the ones not found were actually very small in the BEC. The semantic groups can be seen in Table LXXIV on the next page.

Whilst the PMC did display three of the same semantic groups, it should be noted that the relative amounts were quite different. In the BEC, company/customer relations lexis amounted to over 34% of the sample, whereas in the PMC it was only 17%. Likewise, customer/company lexis was 12% in the BEC, and only 7% in the PMC. Missing from the PMC was the overtly positive lexis noted in the BEC, for instance, dream customer and really good customer, and lexis to describe the type of customer – business customer, professional-type customer. This shows that the actual lexical environment of customer is much richer in the BEC than is presented in the materials of the PMC, and writers have seemingly, in this case, missed an opportunity for teaching the appropriate lexis that normally surrounds the word. The largest semantic groups are represented in the materials, but more lexical accuracy would be required in order to give a full picture of the semantic environment of this word.

TABLE LXXIV: SEMANTIC PROSODIES OF ‘CUSTOMER’ IN THE BEC AND PMC

Left:

BEC PMC

semantic prosody	frequency/ 613 & %	example	semantic prosody	frequency/ 195& %	example
positive	23 – 3.75%	the dream customer a really good customer	size/significance	8 – 4.1%	that big customer was their chief customer loyal to their major customer
negative	4 – 0.65%	a really bad customer
size / number of customers	15 – 2.44%	second largest customer
type of customer (characteristics or line of business)	42 – 6.85%	focus on the business customer the professional-type customer

Right:

BEC PMC

semantic prosody	frequency/ 613 & %	example	semantic prosody	frequency/ 195 & %	example
looking after customers: company -> customer relationship	213 – 34.74%	a customer-driven, enterprise we have to become more customer-focused	company -> customer relationship	35 – 17.94%	customer base customer-conscious customer development
what a customer has or does or is: customer -> company relationship	79 – 12.88%	in response to customer demand ensuing customer dissatisfaction	customer -> company relationship	14 – 7.17%	customer complaints customer expectations customer needs

Product: Product stood out during analysis as a word that showed a limited capacity for semantic prosody, both in the BEC and the PMC. The table can be seen below:

TABLE LXXV: SEMANTIC PROSODIES OF ‘PRODUCT’ IN THE BEC AND PMC

Left:

BEC PMC

semantic prosody	frequency/ 741 & %	example	semantic prosody	frequency/ 526 & %	example
age	33 – 4.45%	entirely new product	age (new only)	62 – 11.78%	new product
			positive	26 – 4.94%	a classic product an excellent product a perfect product

Right:

BEC PMC

semantic prosody	frequency/ 741 & %	example	semantic prosody	frequency/ 526 & %	example
range/choice of products	39 – 5.26%	product range product choice product line	range/choice of products	24 – 4.56%	product lines product range
business activities	60 – 8.09%	product advertising product marketing product creation	business activities	40 – 7.6%	product management product sampling
macro-level demarcation	21 – 2.83%	product segment product area product division

The semantic categories are virtually the same, but, as with customer, there are differences in the relative amounts in these categories. In the PMC there is high stress on new products, 11.78% of the sample, whilst in the BEC, though this category was also present, it represented only 4.45% of the sample. The small category of macro-level demarcation found in the BEC was missing from the PMC. Thus again, both a slightly limited and simplified lexis can be noticed to co-occur with this word in the PMC, though associating semantic groups do closely follow those found in the BEC.

Market: It can be seen from the tables below that the semantic groups associated with market are the same in the BEC and the PMC. Thus, in this case, the word is placed in a very similar semantic environment in both corpora. Having said that, certain differences can be noticed between the two, and these are noted below in Table LXXVI.

TABLE LXXVI: SEMANTIC PROSODIES OF ‘MARKET’ IN THE BEC AND PMC

Left:

BEC PMC

semantic prosody	frequency/ 1,396 & %	example	semantic prosody	frequency/ 946 & %	example
market of what (products/ services)	244 – 17.47%	the airline market the newspaper print market the car market	market of what (products)	113 – 11.94%	the world diamond market the international construction market
where the market is (places)	89 – 6.37%	the European market the French market the Malaysian market	where the market is (places)	136 – 14.37%	North American market Japanese market the New York market
macro-level demarcation	52 – 3.72%	the home market the global market the domestic market	macro-level demarcation	34 – 3.59%	the international market the home market
attributes: positive	53 – 3.79%	a good market a strong market a bull market	attributes: positive	26 – 2.74%	attractive market a highly lucrative market
attributes: negative	19 – 1.36%	a bear market a cut-throat market	attributes: negative	8 – 0.1%	a really difficult market a crowded market

Right:

BEC PMC

semantic prosody	frequency/ 1,396 & %	example	semantic prosody	frequency/ 946 & %	example
a company’s position in relation to others/ the market	146 – 10.45%	market share market position market leadership	a company’s position in relation to others/ the market	50 – 5.28%	market leader market position market share
money/ finance/ economy	81 – 5.8%	market capitalisation market illiquidity market economies	money/ finance/ economy	100 – 10.57%	market economy market forces
macro-level demarcation	33 – 2.36%	market sector market segment	macro-level demarcation	8 – 0.1%	market sector market segment

In the materials there is double the amount of lexis concerned with money/finance than in the BEC (10% to 5%), and there is only half the amount of collocating lexis concerned with a company’s position in the market (5.28% in the PMC to 10.45% in the BEC). There is also a notable lack of lexis related to macro-level demarcation,[207] which was also seen in the case of product above. Frequency of this semantic group was low in both corpora, only 0.1% in the PMC sample and 2.36% in the BEC. The materials were more concerned with where markets were (14.37% in the PMC to 6.37% in the BEC), and less concerned with what kind of market it was (11.94% in the PMC to 17.47% in the BEC). The materials showed slightly fewer positive adjectives than the BEC and also slightly fewer negative adjectives. In all, however, market seems to be fairly well represented in the materials in terms of semantic prosody.

Business: The semantic groups surrounding business were many in both corpora. In general, the same semantic groups were found in both corpora, but there were three more semantic categories in the BEC than in the PMC. Lacking from the PMC, once again, were words concerned with macro-level demarcation (world-wide, overseas, area, sector) and, in this case, this is found for both left and right collocating semantic groups. Also missing from the PMC are links to institutions, organisations and companies found in the BEC. The largest collocating semantic group in the BEC was that of line of business (9.86%), and this was also the biggest in the PMC, but with reduced amount (5.62%). Thus, while the semantic groups found in the PMC more or less match those found in the BEC, they tend to have both fewer examples and to be less lexically diverse in their members. The tables for business are given below:

TABLE LXXVII: SEMANTIC PROSODIES OF ‘BUSINESS’ IN THE BEC AND PMC

Left:

BEC PMC

semantic prosody	frequency/ 2,551& %	example	semantic prosody	frequency/ 1,013 & %	example
where business takes place (place)	56 – 2.19%	Indian business UK business	where business takes place	26 – 2.56%	Middle and Far East business The business of America is business
where business takes place (macro-level demarcation)	123 – 4.82%	international world-wide business overseas business	line/area of business	57 – 5.62%	the steel business the fashion business
line of business	222 – 9.86%	telecoms business hairdressing business contract hire business	nature of business (characteristics)	29 – 2.85%	a family business core business
nature of business (character-istics)	124 – 4.86%	core business family business daily business	money/size of business	22 – 2.17%	big business a billion-dollar business
money/size of business	67 – 2.62%	a high-yield business big business small business	positive	20 – 1.97%	a successful business a flourishing business
positive adjectives	50- 1.96%	successful business sound business strong business	negative	19 – 1.87%	we would simply go out of business the decline in business
negative adjectives	7 – 0.27%	unviablebusiness cut-throat business boring business

Right:

BEC PMC

semantic prosody	frequency/ 2,551& %	example	semantic prosody	frequency/ 1,013 & %	example
people and groups of people	145 – 5.68%	business agents business analyst business controller	business activities	47 – 4.63%	business entertaining business development
business activities	135 – 5.29 %	business administration business analysis business development	people/groups of people	44 – 4.34%	business contacts business travellers
institutions, organisations and companies	129 – 5.05%	business conglomerates business school the business press
macro-level demarcation	89 – 3.48%	business area business sector business segment

9.4.3.1 Discussion

The fact that only five words were analysed and compared between the BEC and the PMC can raise doubts as to the generalisability of the results. However, as the words chosen were of a high ‘keyness’ value and were also relatively frequent in both corpora, it was thought that they would provide a representative sample of how semantic prosody is formed in the PMC. The results of this analysis show that, by and large, materials writers have been able to place the words in appropriate semantic environments. The differences that do occur are concerned with lexical diversity and over- or under-emphasis on certain semantic groups. These differences can be summarised in the table below:

TABLE LXXVIII: DIFFERENCES BETWEEN SEMANTIC PROSODY IN THE PMC AND BEC

Semantic prosody in the PMC	Semantic prosody in the BEC
1. Lexis limited	1. Lexis diverse
2. Focus on money	2. Less focus on money
3. Little mention of macro-level demarcation	3. Focus on macro-level demarcation
4. Simplified lexis	4. More complex lexis
5. Fewer semantic categories	5. More semantic categories

Having looked at semantic prosody in the PMC, the related concept of colligation and the grammar/meaning combinations already studied in the BEC were used to further analyse the same five words. The next section discusses the findings.

9.4.4 Colligation and grammar/meaning combinations in the PMC

Analysis of the five words revealed significant differences in the grammatical patterning presented in the PMC, in comparison to the way in which the words were found to occur in the BEC. Moreover, by a further comparison to COBUILD, the grammar/meaning patterns of the words in the PMC were also found to differ quite widely from those already analysed in the BEC. This section will consider these differences, first looking at colligation and then concluding with a discussion on grammar/meaning differences.

a) Colligation

The differences in the grammatical environments of the words found in the PMC and BEC can be summarised as follows: different use of prepositions, more limited grammatical collocation, less varied use of noun phrases/compound adjectives, and an over-emphasis on personal aspects through the use of possessive pronouns. Each of these will now be looked at in turn.

Different use of prepositions: Prepositional collocation in the PMC was found to be different from that found in the BEC. For example, with the word manager, prepositional usage amounted to 10.6% of the sample in the BEC, whilst the corresponding figure for the PMC was only 4.81%.[208] For the word business, conversely, the PMC stressed in (46 instances) and on (21 instances) business. The corresponding figures for the BEC showed much less use of these combinations (in business 48 instancesand on business 15 instances), bearing in mind the BEC is twice the size of the PMC. Moreover, this prepositional usage resulted in grammar/meaning stereotyping.[209] Even though on business was found 15 times in the BEC, not one of its instances corresponded to the meaning given in the PMC for the combination. In the PMC, almost all the examples were of the I’m here on business variety (19/21 instances). In the BEC, on was simply used as a grammatical collocation to words preceding business, for example, concentrate on business, doing a talk on business, burdens placed on business. Materials writers, understandably, need to stress the prepositions that can be used with business as it is an important word, but, in over-stressing one aspect of it, they have given a limited picture of actual usage.

More concentration on limited aspects of grammatical collocation: There seemed to be a much simplified grammar being presented in the PMC. The word business, for example, co-occurred with forms of do 59 times in the PMC (5.81% of the sample) whereas it only occurred 41 times (1.6% of the sample in the BEC). A preoccupation in the PMC with the grammatical collocation of do and business is, therefore, clearly apparent. Again, as with on and in business, this is an understandable feature as students commonly mistakenly use make with business. The result, is, however, that limited grammatical patterning is presented to students.

Lack of or much less varied use of noun phrases/compound adjectives: It was noted previously that some words had a strong tendency to be formed into noun groups/compound adjectives. Customer, for example, displayed this tendency very clearly in the BEC. In the PMC, some words showed similar tendencies but not to the same degree, whilst other words were completely lacking in these typical features. In the case of customer, the tendency was recognised, but was used to a much lesser degree. Interestingly, considering previous discussion on the use of problem situations in the materials, the collocation customer complaints appeared 4 times in the PMC out of a total sample size of 195, but occurred only 3 times in the BEC out of 613 instances. Another example, market, formed into a high number of compound nouns in the BEC, e.g. market share, market leader and market segment, but the examples found in the PMC were both fewer and also much less varied, with most instances covered by three collocations – market share, market leader/ship, market research. A further feature noted in the BEC about market – its strong tendency to be post-modified into noun phrases (e.g. market-driven solutions) – was almost totally lacking in the PMC.

Over-emphasis on personal aspects: This has been noted in terms of lexical usage in the PMC – here it can also be seen on a grammatical level, with the high use of possessive pronouns in the case of certain words. This could be slight, for example, with customer (4.24% in the BEC to 6.15% in the PMC), or greater, as with product, which was found in collocation with possessive pronouns 26 times in the BEC – 3.5% of the sample – and 56 times in the PMC – 10.64% of the sample. It was found that not all words displayed these differences in usage, and further study would be needed to clarify this point. These differences shown in terms of colligation obviously had a bearing on the grammar/meaning combinations found in the PMC.

b) Grammar/meaning

Comparison of the COBUILD classifications for the same five words between the PMC and the BEC showed marked differences between both the limitation and the most common meanings of the words used. The table below compares the limitation, the order of the most common meanings, and the identification of new meanings found in the words for the PMC and the BEC:

TABLE LXXIX: COMPARISON OF FIVE WORDS PMC/BEC

PMC Word	Limited use	Change of order	New meaning	BEC Word	Limited use	Change of order	New meaning
manager	Yes: 1/3	–		manager	Yes: 2/3	–	–
customer	Yes: 1/2	–		customer	Yes: 1/2	–	–
product	Yes: 1/2	–	Yes	product	–	–	Yes
market	Yes: 11/12	3-2-11-4-7		market	Yes: 9/12	3-2-7-5-11 no No.1	–
business	Yes: 13/22	1-3-5-2-4		business	Yes 15/22	1-3-5-2-12	Yes

Three of the words – manager, product and business – showed that their use was more limited in the PMC than in the BEC, i.e. less senses of the word were used in the PMC. Conversely, there were two more senses of market found in the PMC than in the BEC. Three words, product, market and business, were especially interesting in their comparative use between the two corpora, and these will now be discussed in more detail.

Product: The differences between the use of product in the BEC and its use in general English were noted earlier. The differences in the use of product between the PMC and the BEC are just as marked. In the BEC, two senses of the word were found – something produced and sold in large quantities and product of a situation – where only one was used in the PMC – something produced and sold in large quantities. Moreover, the key finding of product in the BEC, i.e. its usage as an uncount noun (amounting to 14.3% of the sample) was almost totally missing from the PMC. In the PMC this patterning was only found in two instances, both from the same line of the same book – Presenting in English (Powell 1996). This means that an important usage of the word is, in practice, completely missing from the materials that should, in fact, be stressing it in order to give students an accurate picture of its usage.

Market: Here, there was a slight difference between the materials and the BEC. In the PMC, the sense of market referring to the amount sold in a market[210] (COBUILD Sense 3) came to 34% of all examples. In the BEC this was used in 37% of the sample. A further difference, also slight, was the fact that in the BEC the most common COBUILD sense of market (i.e. a place to go and buy goods) was absent, but was found in the PMC – though only one instance of it was found (market town).

Business: The semantic categories of entertaining and travel were noted earlier as being over-stressed in the PMC. In the case of business, this stress also filtered down to a grammatical level. COBUILD Sense 4 (what you do for your job and not for pleasure), which has a common pattern of on-noun, was used 61 times in the PMC (6.01% of the sample), whilst in the BEC, it was only found 8 times (0.31% of the sample). Examples of this grammatical stereotyping from the PMC are shown here:

Thus, the concentration on travel and the distinction between business and pleasure also has grammatical consequences. So too does the stress in the materials on problem situations. The instances of COBUILD Sense 20 (go out of business) in the PMC are more than three times more frequent than in the BEC (0.31% in the BEC to 1.08% in the PMC). The materials writers views of business also come to the fore in well-known collocations: business card/s is found 11 times in the PMC (1.08% of the sample) compared to only two instances in the BEC. Likewise, business class is twice as frequent in the PMC as in the BEC (8-4). Admittedly, these last figures are all very small, but arguably they all work towards creating a picture of Business English that can be found in the materials that is not matched in the BEC.

To summarise the findings of this section it can be seen that:

1. The grammatical patterning found in the PMC often differed from that found in the BEC.

2. Certain grammatical features thought important by writers are over-represented to the detriment of others, for example, personal/possessive pronouns.

3. Lexical meaning was limited in the PMC, as certain senses/meanings of the lexis were stressed over others, not matching what was found in real-life business lexis.

4. Exposure to limited meanings means exposure to limited grammar and colligational patterning. It can be seen that certain meanings have very common grammatical patterns, e.g. on business, out of business. When some meanings are chosen over others, the consequence is that students receive restricted and, therefore, skewed exposure – too much exposure to some lexis and too little to others.

5. Finally, grammatical stereotyping in the PMC attaches one meaning to one pattern – as with on business. The lexis in the BEC displays much broader grammatical patterning that is not represented in the materials.

9.4.5 Word clusters in the PMC

Analysis here was limited to looking at only 3-word clusters, though expansions of these chosen 3-word clusters, going up to 5-word clusters, were also utilised. The clusters were investigated both in terms of frequency and keyness by computing lists using WordSmith 3, and also at a micro-level by examining clusters formed from the 5 words in the PMC singled out for analysis.

In all, the following lists were computed for comparison:

1. The most frequent 3-word clusters in the PMC.

2. The most frequent 3-word clusters in the BEC for comparison.

3. The ‘key’ 3-word clusters (BNC as reference corpus).

4. The ‘key’ 3-word clusters (BEC as reference corpus).

Fuller lists of PMC 3-word clusters can be found on the CD ROM attached to the back page of this thesis, but the top 15 of each of these lists are shown on the next page:

Fig.61 Most frequent 3-word clusters PMC Most frequent 3-word clusters BEC

Fig.62 PMC-> (BNC)Key words PMC -> (BEC) key words

9.4.5.1 Clusters at a macro-level in the PMC

The lists computed of 3-word clusters confirms the findings of analysis at the level of single words. In fact, it can be said that the lexical characteristics of the PMC become even more apparent at the 3-word level, and these are now discussed below.

a) Polite personal contact: The PMC is full of clusters that are connected to both politeness and personal contact, and combinations of these two characteristics. This is true both in terms of frequency and also in terms of keyness, independent of which corpus is being used as a reference corpus. Just looking at the most frequent 3-word clusters of the PMC shown above in Fig. 61, the most frequent, I’d like to, forms the functional stem of a large number of requests coming from telephoning (I’d like to speak to) and meetings (I’d like to draw your attention…).

The fourth most frequent 3-word cluster in the PMC, would you like, shows similar tendencies:

These two examples, therefore, show a preoccupation in the materials with politeness and interpersonal relationships. They also point to the second notable aspect discussed below, that of high genre- and situation-specificity.

b) Clusters in the PMC are highly genre- and situation-specific: There was a difference in the type of 3-word clusters that were most frequently used in the PMC and BEC. In the BEC it was found that the most frequent shorter 3-word clusters were fairly genre-independent, though they could be defined in terms of the written/spoken and the doing business/talking about business axes. In the PMC, there was a high concentration of clusters that are mainly used in a small number of business situations. The would you like example above shows usage in entertaining situations (would you like: a drink, coffee, to see, to go to, to go out), on the phone (to hold), and meetings (to begin, to start). In addition to this type of cluster, there were also very frequent clusters that were used almost entirely in one genre. For example, the cluster thank you for is found almost exclusively in the genre of business letters:

c) There is a concentration in the PMC on a limited number of set phrases: The trends noted above were reinforced by further analysis, with a limited number of phrases being repeated. For example, the phrase get down to business occurred 14 times in the PMC (in a total of 7 books) and not once in the BEC, despite the inclusion of a large number of meetings. The examples from the PMC are shown below:

What must be pointed out here is not the fact that the clusters are being used incorrectly, or necessarily in a different way from in the BEC or the BNC, but that they only represent a very limited view of business through a restricted lexis. The materials writers view of business comes across very strongly again – politeness, interpersonal contact, correspondence, travel, entertaining and food.

9.4.5.2 Clusters at a micro-level in the PMC

Differences in 3-word clusters were also found as part of the analysis of the original 50 words in the BEC. The five words under analysis from the PMC, compared to their use in the BEC, showed certain differences in terms of the kind of clusters formed. A comparison of the clusters is shown in Table LXXX below.

This small sample shows that it is possible to identify certain tendencies in the clusters found in the PMC. It was noted when discussing how the word business was used in the PMC that there was a focus in the materials on limited aspects of its grammatical collocation. This can be seen clearly here. The most frequent cluster is the same, of the business, but the other two (to do business and down to business) confirm the narrow focus of the materials noted earlier. The same tendency can be observed with the word manager. This word showed a tendency in the BEC for grammatical collocation with prepositions, notably of. In the PMC this is replaced by job titles (export sales manager, production manager). The word product, again, shows the same tendency, grammatical collocation being replaced with concrete items – this time with a focus on new products.

TABLE LXXX: THE MOST FREQUENT 3-WORD CLUSTERS BEC AND PMC

BEC 3-word clusters	Freq	Freq	PMC 3-word clusters
manager
manager of the	12	12	exports sales manager
the manager of	8	12	the production manager
general manager of	6	11	software engineering manager
customer
to our customers	13	8	to the customer
to the customers	8	7	the customer is
of its customers	7	5	if the customer
product
of the product	29	19	a new product
the product is	16	14	the new product
a new product	14	12	of the product
market
in the market	44	43	on the market
the stock market	39	38	in the market
of the market	38	20	of the market
business
of the business	85	34	of the business
in the business	63	16	to do business
to the business	31	15	down to business

9.4.6 The PMC: Conclusions

At the beginning of this whole section on the PMC, four questions were asked, in addition to the statement of Hypothesis 2 and the main research question. These questions can all now be answered. Hypothesis 2 stated that the lexis used in published Business English materials was significantly different from that found in real-life business situations. It has been found that there are significant differences between the language used in published materials and in real-life. Whilst much of the same language is used in both, there are significant differences in terms of frequency, usage and complexity. The materials present a picture of the world of business that is restricted in terms of its lexis, grammatical patterning and word clustering. In order to summarise these differences, the four questions asked earlier are reiterated and answered.

1. How do the PMC key words define the lexical world of business and how does this definition compare to that shown in the BEC?

The world of Business English presented in the PMC is one of a finite number of contextualised business situations often framed by reference to problems. It is a world of meetings, presentations, travel, entertaining and food. It is a world where both negative and positive feelings can be expressed, a world where personal and interpersonal lexis is common, and a world where politeness is very highly valued. The focus of this world is on tangible objects and reference to states and qualities is limited. In short, it is the world of business viewed though the eyes of teachers, where business matters are expressed more through noun/verbs than verbs, in language that can be emotive and, more seriously, lacking in the substance of business. The results gained from the PMC reflect, to some extent, the categorisations of published Business English materials made by St John (1996:9-14). St John’s first category of materials – materials for business communication – focused on general, non-business, communication skills needed to function in different walks of life, and it is just this category of language that shows up clearly in the PMC.

The reference to the substance of business above is meant with regard to the relative lack of both pure business, sub-business and money/finance-related lexis found in the PMC. This must be considered a very significant lack. Although many business-related words were computed as being key words when compared to the BNC, they were negative key words when compared to the BEC (though this finding does not necessarily prove any lacks). Much of the lexis that was rejected by the BEC as being non-business – personal and interpersonal lexis – was heavily stressed in the PMC and where the BEC had a high incidence of positive lexis, the PMC equally distributed positive and negative. In summary, the lexis found in the PMC was simpler, more concrete, less varied, more polite and much more focused on human interaction than that found in the BEC. These findings confirm earlier work on the lacks of both Business English and other EFL materials noted in Chapter 3. The work of Kennedy (1987), Holmes (1988), Williams (1988), Scotton & Bernsten (1988) and Boxer & Pickering (1995), for example, found the same concentration on politeness,[211] a focus on the explicit over the implicit and a restricted and simplified language that did not match real life situations.

2. What semantic prosodies were found in the PMC and do they match or differ from those found in the BEC?

Similar findings were made with analysis of semantic prosody in the PMC, though analysis was limited to five words only. The materials were able to form very similar semantic groupings with which the lexis associated, which must be considered a big plus for the materials. However, although similar semantic groups were found, the number of semantic groups found was less, and the lexis they contained was much less varied. Thus, again, a more simplified view of lexis was presented, resulting in a narrower view of the lexical world of business.

3. What colligational and grammar/meaning patterns were identified in the PMC and how do they compare to those found in the BEC?

The representation of semantic prosody in the materials could be described as relatively good. However, colligational patterning and grammar/meaning distinctions were not as accurate. Colligationally, there were differences in terms of prepositional usage, a concentration on limited aspects of grammatical collocation, and a lack, or much less varied use, of the lexis when it was used to form noun phrases. Also noted was an emphasis on personal aspects through high usage of personal/possessive pronouns. In terms of grammar/meaning categories as defined by COBUILD, there was an over-representation of some meanings over others that did not match those found in the BEC. Additionally, in the case of product, a sense noted in the BEC was almost completely missing. Thus, by focusing on some meanings rather than others, not only was a distorted picture of word meanings given, but also typical grammatical patternings were overlooked as patterning is commonly tied to meaning.

4. How do the clusters found in the PMC compare to those in the BEC?

All the above points can be applied to the clusters found in the PMC. An over-emphasis on politeness, social situations, travel and a high incidence of genre-specific clusters were in evidence. The latter was true even in the shorter, 3-word clusters, which differed markedly from those found in the BEC, where the clusters were much more genre independent.

9.4.6.1 The next section

Analysis of the BEC showed the basic lexical characteristics of Business English. Analysis of the PMC illustrated the way in which lexis in published materials differs from that found in the BEC. Thus far, the discussion has centred on purely linguistic aspects. Yet, equally important must be the pedagogical consequences of this study. If the lexis in the materials and real life differ, is this difference acceptable, or must it be rectified in future materials? The next section will address pedagogical issues brought about by the analysis of the BEC and the PMC.

9.5 Pedagogical Issues

After the density of the analysis of the BEC and the PMC, this section is short and essentially practical in nature. Discussion here focuses on how the results of this thesis can affect the creation and design of Business English teaching materials. Much has been written in the literature about the use of corpora in the classroom, teaching techniques that have been developed using corpora and concordancing, and the effect corpus-based learning can have on students (Johns 1988, Ma 1993a, Murison-Bowie 1996, Owen 1996, Wichmann et al. 1997). Whilst very important, this aspect will not be comprehensively covered here, as these authors, and others, have provided a detailed picture of the area. Rather, this part of the chapter primarily takes the form of recommendations for future materials, these recommendations coming as a direct result of work done both with the BEC and the PMC. It has been stated from the outset that this work is essentially pedagogic in purpose, and so not only recommendations for future materials are offered, but also concrete examples of them (example materials can be found in Appendix 11 in Vol. II). These materials, of course, only present a small sample of what needs to be done in the future. It is hoped that the materials produced for this thesis will be the first stepping stone to a large amount of Business English materials that can be made available for students to take advantage of. In the section that follows, eight main recommendations are made with regard to the development of Business English teaching materials:

materials creation should be corpus-based
Business English materials should contain Business English
sub-business language needs to be stressed
semantic prosody needs to be made widely known and be explicitly taught
lexis should be seen in its typical grammatical setting
students should know that words are not evenly distributed
there should be a greater focus on word clusters
associate words should be more recognised

9.5.1 Materials creation should be corpus-based

One overriding theme has been seen throughout this thesis: the contrast between teachers’/writers’ intuition and the results of empirical investigation. The work of Williams (1988), Kennedy (1987), Holmes (1988), Pickard (1992) and Ljung (1990), to mention but a small number of the studies comparing ‘real-life’ language to published materials, all point to the inaccuracy of materials in a variety of aspects. The incident described in the following extract gives further indication of the dangers of intuition:[212]

One of my students recently wrote onto an OHT: ‘There has been a boom in the number of patients dying when taking this drug…’

I corrected him saying that there is something wrong – the combination of boom with all its positive connotations and patients dying creates a contradictory message. Boom, I thought, has a very strong positive semantic prosody related to good times and a surfeit of money.

After correcting the student I consulted the BEC and some of the examples found are shown below:

Here, it can clearly be seen that boom has a mixed prosody: bankruptcy can boom as well as sales, businesses can enjoy a boom and booms can also be dangerously inflationary. This incident spells out two clear messages: firstly that words do in fact tend to have semantic prosodies and secondly, that these prosodies are not necessarily what teachers may intuitively think they are.

Intuition, then, on the part of teachers or materials writers, can be misleading, and the use of corpora for materials creation offers more certainty in terms of knowing how language actually behaves. Yet of central importance here is the corpus used. Lewis (2000) pointed out ‘the single most essential thing about developing or using a corpus is that it must be designed for a particular purpose’ (Lewis 2000:193). This means that not only the target group of students must be clear, but also the purposes to which they need to put the language. Thus, as Leech notes ‘the best kind of corpus would be one corresponding to the target language behaviour of the learners concerned’ (Leech 1997:19). This corpus does not necessarily have to be a large one, but it needs to be chosen with care. Tribble’s (1998, forthcoming) work with what he termed ‘exemplar texts’ in micro-corpora, noted previously, suggests that this can be done successfully (for example, his corpus of web leaflets promoting MA courses is only 13,216 words). He states ‘By studying an examplar corpus in contrast to a reference corpus it is possible to gain an understanding of the texts which are similar to and different from the texts the learners want to write’ (Tribble, forthcoming). Tribble’s focus here is obviously on genre, but the same principle of using a larger reference corpus to determine the key characteristics of the lexis of a smaller corpus can apply to the development of a variety of materials, not just those concerned with genre.

In contrast to the pro-corpora writers, there has been some debate as to the efficacy of corpus-based teaching materials and the use of concordancing as a source of knowledge for teachers, notably with regard to the COBUILD project and the materials that it has engendered. Owen (1996) pointed to the diverse and often inaccurate language that can be found in a corpus, and suggested caution in using corpus evidence in teaching: ‘Even if teachers had the time to check every prescription they want to make, the corpus would not relieve them of the burden of using their intuition’ (Owen 1996:224). It is true that corpora do often consist of language that can be considered non-standard, or even ‘wrong’ from the perspective of the student, but the accumulation of data showing how language actually works, as opposed to how it was thought to work, cannot be ignored. The difference between the creation of materials purely from intuition, or as a result of a combination of corpus evidence and intuition, can be wide, and this thesis has shown the clear contrast between real-life business lexis and the lexis found in published materials.

There must now be a gradual transition in the way Business English materials are created, with corpora playing a much larger role. It is hoped that the availability of the BEC to other researchers in the future will play a part in this change. In this way Business English materials can contain much more accurate Business English.

9.5.2. Business English materials should contain Business English

The lexis used as the basis for study in this thesis has been gained by finding statistically ‘key’ words. It therefore differs from most earlier works that have focused on pure frequency of occurrence. Leech (1997: 15) noted that the central philosophy behind the COBUILD project, discussed above, was one of frequency, where frequency data directly affected the writing of materials. Willis (1990:46) showed a table pointing out that the 700 most frequent words of English make up 70% of English text, the 1,500 most frequent words make up 76% and the 2,500 most frequent words constitute 80% of the text. Therefore, for the teaching of general English it would seem to make sense to concentrate on these words to give learners the best possibility of communicating effectively. Yet when considering a ‘specialist’ language, many of the most frequent words in Business English are the same as in general English. The use of key words, that is, unusual frequency as opposed to pure frequency in this thesis, has filtered out the business lexis from the general. The key word analysis of the BEC has shown that the business world is marked out lexically from general English and it is into this lexical world that students should be immersed. To some extent this presents a restricted view of lexis, similar to the vocabulary limitation movement connected to Palmer, Ogden and West in the first half of the twentieth century. But whereas the vocabulary limitation movement saw the words as an end in themselves, and wanted to present a simplified language, the approach here sees the words, i.e. key words, only as a starting point. This allows for the investigation of polysemy, homophony, collocation, clustering, colligation and association and places the lexis in a rich business-related environment.

It has been suggested that the key words computed in this thesis represent core business lexis. The use of this lexis as the basis for teaching materials presupposes the accuracy of the corpus, the reference corpus and the statistical procedures used for computing them. The validity of these aspects has been defended elsewhere in the thesis, but it must be stressed that the key words do not and could not represent the entire picture of Business English. What they do give is a manageable number of words, just under 2,000, that can be used as a first resource for materials writers. The key words have been purposely divided into semantic groups to afford both teachers and students easy access to the lexis.[213] It has been noted that these semantic groups recur throughout and also figure significantly in terms of semantic prosody. This provides a unity and helps define the world of business as it is seen by business people, as opposed to materials writers. This world can be seen summarised again below:

TABLE LXXXI: SUMMARY OF THE LEXICAL WORLD OF BUSINESS

Main semantic groups		Main features of the lexis
business people	business events	business-related
money	quantities & measures	positive
communication	business activities	shallow
technology	business objects	dynamic public verbs
business places		non-emotive
running a business		limited, mainly positive states & qualities
companies/institutions		impersonal adjectives
time		concrete

Several points logically follow on from the above:

1. The world of business lexically demarcated above needs to be an explicit part of teaching materials: Students can be given a clear idea of the lexical world into which they are venturing, in the same way that travellers have a map to guide them, and the semantic categories help present a concrete picture of business language. Presenting Business English students with information in a concrete form can have a positive influence on the teaching situation, due to the special needs of business learners. As Lees noted already in the early 1980s, ‘business people form a species apart, and require a special diet’ (Lees 1984:133). Part of this diet is the need for immediate motivation and the perceived relevance of the materials they use. The students can, therefore, be explicitly told what the key semantic groups are and then be systematically initiated into the lexis of each group.

The world of business lexis

Semantic Semantic

groups groups

You are here

Fig. 63 A lexical map of Business English for students

The students are thus given a semantic map of Business English lexis. The students should also be told, however, that this map is incomplete. It shows the main roads and larger features of the area, but the detail is not fine enough to show the minor roads and smaller special features of the landscape. Thus, more work would need to be done to cover the specialist lexis of a student’s given field, which has largely (intentionally) been filtered out from the key words.

2. There is an emphasis on positive states and qualities in the key words of business: Students need to be aware of this positive focus and materials need to somehow pay less attention to problem situations than they do at present. This is difficult, as writers have often presented Business English tasks in terms of problem-solving situations, and this may well be psychologically motivating. However, at the same time, the amount of negative lexis the students are exposed to in the published materials outweighs its importance. It is possible to compromise. Students could be explicitly taught the positive lexis and be told of its frequency and importance in Business English. No harm can come to students by teaching negative language as long as they know how and when to use it, and that it is used less in Business English than in everyday life.

3. A clarification of the politeness-positive language continuum: Both previous studies and this one have shown that published materials are over-polite and focus a lot on the lexis of politeness strategies. This over-emphasis on politeness is not new, and was reported as early as the 1980s (Williams 1988). What is new here is that fact that the BEC shows that business language, whilst not necessarily over-polite – though it is rarely impolite – clearly veers overall towards the positive. In the PMC there is a more even distribution of positive and negative language, and, as has been mentioned, there is a focus on problems. There thus needs to be less concentration on being polite, and more concentration on less personal lexis that indicates positive qualities – words such as effective, available, dynamic and innovative.

4. Materials derived from the key words need to be suitable for all levels: The problems of using concordancing in the classroom has been widely reported (Owen 1996, Wichmann et al. 1997). The materials made from the BEC need to be suitable for all levels of students’ language ability. Aston (1997) quite rightly points out that ‘Learners may need practice in identifying patterns of collocation, colligation, connotation and discourse structuring; and at the same time to appreciate the relevance of noting such regularities in their own learning’ (Aston 1997:60). There should therefore be a balance of materials: more concrete, straightforward materials for lower ability levels who are not as easily able to cope with deductive, explorative concordance work, and more concordance-based, deductive materials for the more advanced: ‘Corpus-based activities need to be designed bearing in mind the need for learners to acquire the ability to exploit these resources’ (Aston 1997:61). For an example of materials developed from the key words see the Appendix 11 in Vol. II (Exercise 1, p.892 – Exploiting key business words – nouns).

5. Business materials should not, however, be lexically over-restricted. A final word of warning: the fact that a given word became a negative key word in the BEC does not necessarily mean that it is not used in business. It may mean it is used, but is simply used less. Therefore, whilst semantic boundaries can be placed around the business world, it has been noted that these boundaries are very much open to interpretation, and teaching should not be limited only to the words presented in this thesis.

9.5.3 Sub-business language needs to be stressed

In their review of Business Objectives, Ayers & Van Huyssteen (1996) observed that ‘it seems that a grammatical syllabus has been drawn up, very much like for any General English course book, and a list of ‘business situations’ ‘matched’ to those structures’ (Ayres & Van Huyssteen 1996:74). This observation is backed up by findings in the PMC – a focus on certain grammatical patterns and a stylised and limited view of Business English lexis. There needs, therefore, in line with Sinclair’s (1991) idiom principle, to be a stronger lexical focus in the materials – so that the focus of learning is on the generative power of business-related lexis, as opposed to a slightly disguised grammatical syllabus with business words included. There also needs to be a greater stress in the materials on what has been termed in this thesis as ‘sub-business lexis’, a borrowed term from sub-technical vocabulary. Lewis (2000) explains its importance:

This sub-technical vocabulary lies between general English and the technical vocabulary of a particular specialism, and is of great importance to ESP learners, as it is precisely this language which they need to communicate about their specialism to non-specialists, such as patients, supplier or customers. (Lewis 2000:195)

Sub-technical vocabulary has been defined as words which are context independent and have a high frequency across disciplines (Cowan 1974), and words that have one or more general meanings and take on different meanings in certain contexts (Trimble 1985). Lewis’s (2000) definition suggests both words and phrases that operate in the background of a specialty. Thus students may know cardio-vascular, but may not know collocations such as straighten your arm (Lewis 2000:195). It is just these types of words and phrases in the background of business that need to be explicitly taught to students. In the BEC they have been categorised as words concerned with running a business. Below are examples of these words taken from the noun/verb category of the BEC positive key words:

plan, forecast, control, issue, value, design, request, venture, ship, address, store, flow, book, handle

None of these words are ‘pure’ business-related words, but on an intuitive level they can all fit into the lexis surrounding a business. They are words that have one meaning in general English and another meaning, or at least another application, in Business English. These example words could be exploited for their verbal and nominal meanings and collocative- and cluster-based work could be generated. Thus, students learn the vocabulary in a business setting, with language fanning out in ever greater circles from the one original word.

Example exploitation: Nominal use of plan, for example, could be exploited by looking at what kind of plans are common in business:

The data above shows business plan as the overwhelmingly most common cluster, but other examples can be found (by concordance) related to concrete nouns: sales plan, repayment plan, tax plan; terms concerning time: longer-term, medium term plan and positive adjectives: detailed, realistic, well-documented plan. Likewise, verbs preceding plan as a noun: put together, compose, develop, reject, enter into and prepare a plan. Materials could then show that these phrases can lead on to longer phrases: put together a business plan, reject a finance plan and so on. Additionally, looking at plan as a verb, the BEC can show what business people plan:

These kind of exercises focusing on collocations are not new (see for example, Lewis & Willberg’s Business English 1990), but the difference here is that the BEC is able to generate authentic business examples, and the teacher’s/writer’s job then is mainly to present the information in such a way as students will best learn, rather than create the examples themselves.

9.5.4. Semantic prosody needs to be made widely known and explicitly taught

The concept of semantic prosody can be of great value to both teachers and students alike and should be explicitly taught. Analysis of the PMC has shown that, whilst similar semantic prosodies could be found for words, the prosodies were less in number, less lexically rich, and most importantly, the prosodies were not specifically taught – semantic prosodies are thus presented implicitly in the PMC rather than explicitly. Hoey (2000) and Stubbs (1995) both point to the need to make the concept of semantic prosody known as ‘You would scan the pages of most of my language coursebooks in vain for the slightest hint that words have semantic prosodies’ (Hoey 2000:233). Hoey, in the same article, points to one central importance of semantic prosody by saying that semantic prosodies

… are potentially powerful generalisations for the language learner. It is no longer necessary to learn collocations as individual combinations…. Instead, what the learner needs to do is to learn the word in combination with an absolutely typical representative of the prosody as long as (s)he also knows that it IS typical.

Hoey (2000:233)

This typicality means that the typical prosody of a word is divided into positive or negative connotations, or that a word is commonly tied to a specific semantic group or set. This has many consequences for the design of materials. By making semantic prosody an overt part of Business English teaching materials, and by using it in conjunction with evidence gained from the BEC, several significant improvements can be made in future materials (laid out 1-4 below). It should be stressed here once again that analysis by semantic prosody is a concept only made possible by access to a computerised corpus. In order to study semantic prosody in Business English, therefore, a Business English corpus is an absolute requirement. The BEC is able to provide this data.

1. Words can now be taught with lexical and collocational accuracy: Hoey’s comment above about limiting the number of collocates necessary to teach is valid and may well be of use in certain teaching situations. However, semantic prosody also gives the opportunity of presenting the rich and diverse lexical environment surrounding Business English words. It can also do this accurately. Some examples of this are shown below using the words provide and competitive:

Provide: This word has a very positive semantic prosody and students can be taught this explicitly so as to prepare them for possible collocates. The examples below show provide collocating with a wide range of positive co-text.

After the positivity comes a whole range of business-related items that the positivity is attached to, for example, third party certification services, operational service, and here students could add their own business activities (as done by Willberg & Lewis 1990).

Competitive: Competitive had two semantic prosodies that would be useful to be noted by students – firstly its common co-association with extremes (15% of the sample):

and secondly, with atmosphere (4% of the sample):

This kind of lexis needs to be explicitly presented to students as being typical of how the word behaves in the business environment.

Students can be taught that some words have unique semantic prosodies, some words share the same prosodies, and some words exhibit both tendencies. Sale, for example, has a unique prosody connected to availability, whereas the word package, whilst having a unique prosody connected to computers, also shares a prosody related to finance with other words such as merger, market and finance. It is important, therefore, that students know that certain semantic themes run throughout Business English, but also which semantic groups individual words may be connected to. An example of materials making use of semantic prosody can be found in Appendix 11 in Vol. II, p.900 (Exercise 4 – Exploiting semantic prosody).

2. Lesser known collocates can be easily included in materials: Continuing the theme of lexical richness, semantic prosody also allows less obvious collocates to come into view. Two clear examples can be given here. With the word employee, a semantic prosody was found with employee benefits, and the collocation employee benefits is, perhaps, quite well known. However, falling into the same category were other collocates such as employee health care and employee share ownership, which are perhaps not as widely known. A similar example was noted with the word strategic, where a wide range of collocates all fitted with the overall semantic set of aiming for the future. This ranged from the common collocation strategic plan, to the lesser known strategic direction. Thus, semantic prosody provides a richness of lexis for students to see, yet at the same time it displays the lexis within a semantic framework that allows for simple categorisation and therefore, hopefully, easier recall.

3. Semantic prosody can show students how collocates become more fixed in the Business English environment: It was found that certain common words, such as big, which in general English have very non-fixed collocational partners, became more fixed in the business environment. Additionally, analysis of words in both the BEC and the BNC showed that all the words became more restricted in collocational partnerships. This can possibly aid lower level students, as these words can be taught in collocation with their most typical, and more limited, business-related prosodic partners. It should be stressed here that this fixedness is relative – big, for example, still collocates widely, simply not as widely as in general English.

4. Semantic prosody can show students how business people think: The way in which words behave is a direct result of the purpose they are put to in the language, and consequently how they are semantically perceived by the people who use them. Semantic prosody enables these thought patterns to be made clearer. For example, the phrase market share was noted to associate with a prosody of obtaining and improving in 43% of all instances of the word:

In contrast, a semantic prosody of losing market share amounted to only 7% of the sample. This indicates that business people are more likely to discuss market share in terms of gaining rather than losing it. Likewise, a semantic prosody of limitation/control (6% of the sample) with the word export, denoted the need for export control:

A further interesting example was found with the word boss. In the BEC, no positive adjectives were found relating to boss, but there was a small group of negative adjectives (e.g. meanest, old-fashioned) and some negative co-text (e.g. the boss is a neanderthal). This compares with the word manager, that did display both positive collocates and co-text. This indicates that business people may refer to their bosses unflatteringly, but the more formal term, manager, takes on a more neutral or positive role. By explicitly showing the lexical environment of words in this way, materials can prepare the students for the actual business world they may need to work in, or give them more information about areas that they already do work in.

9.5.5 Lexis should be seen in its typical grammatical setting

Every bit as important as portraying language in its correct semantic setting is showing it in its correct grammatical environment. In fact, as has been discussed at length in this thesis, the two cannot be satisfactorily separated. Yet it is possible to talk about language from both standpoints. The lexical standpoint was covered by discussion of key words and semantic prosody, and in this section emphasis is laid on the grammatical aspects of the words.

1. Words should be presented with colligational accuracy: It was found that present Business English materials showed several lacks in terms of presenting the grammatical environments of words. One of the key lacks was inaccurate or limited presentation of the grammatical patterning of words most commonly found in the BEC. Use of the BEC enables a more accurate presentation to students of the patterning found. For example, the fact that the word production is often followed by plural nouns:

or that market is often formed into a post-modified adjective fitting into noun phrases;

and that performance is often part of a three-word noun group co-joined by and:

This kind of detail was found to be lacking in the PMC and one notable improvement that can be made to materials is to take this typical grammatical patterning much more into account. This, too, cannot be done satisfactorily without reference to a relevant corpus. For an example of materials focusing on colligation, see Appendix 11 in Vol II, p.893 (Exercise 2 – Exploiting typical grammatical patterning).

2. Unique colligational patterns should be presented, but not over-represented: It has been seen that Business English lexis often uses one sense of a word more than another, and that word sense is tied to grammatical patterning. For example, the most common sense of market in general English (a market place where you go to buy goods – count noun) was absent from the BEC. Instead, the most common meaning found was market in the sense of how much product is sold, with the pattern of the-noun (we control 60% of the market). Thus, meanings are more specific and limited, and as a consequence, grammatical patterning is not as broad as in general English. This has distinct pedagogical implications in that some meanings and grammatical patterns should be taught more than others. This has been done in the PMC, where certain grammatical patterns were found to be stressed (e.g. on business, doing business), whilst little or no overt attention was paid to others (e.g. patterning discussed earlier in relation to interest rates, high and low). This is a natural consequence of the teacher’s/writer’s job – the singling out of key grammatical features for students to take note of and learn. Thus it is accepted that a pedagogical rather than general grammar should be given to students. However, when singling a feature out for repeated practice, one has to be certain of its value.

It is argued here that the range of grammatical patterns presented to students needs to be broadened and a more realistic grammatical picture of Business English explicitly presented. Moreover, it is not enough just to have correct examples implicit in the exercises – students should be told what is typical for a given word. For example, the positive prosody/grammatical patterning found with the word communication – positive adjective + noun + noun (group):

Also to be noted here is the genre-specificity of this pattern – all instances came from job advertisements. This broadening should be done, however, whilst still paying regard to the ability level and interests of the individual students.

3. Grammar/meaning contrasts between Business English and general English need to be explicitly laid out: Discussion on sub-technical or sub-business vocabulary above noted the importance of this kind of language to students. The meaning/grammar distinctions made in COBUILD and this thesis enable both business-specific meanings and their typical grammatical formations to be a greater part of the students’ understanding of business lexis. Thus students could be taught, to give a very simple example, that in general English the word partner has one main meaning (life or sexual partner) but in business the meaning is altered (people who share ownership in a business). They could also be taught that in Business English a further sense is found where partner is used in the pattern verb-with-noun:

Grammar and meaning go hand in hand, which leads to the final point.

4. Students need to realise that lexis and grammar are not separate entities: All the work done on the BEC confirms previous studies showing that grammar and meaning are interrelated concepts, and consequently they do not need to be taught separately. Rather, students need to be taught how they interrelate. This is true as much for business lexis as for more general words. Work done in COBUILD enables, for example, the word high to be taught both in terms of its different meanings and in terms of the grammatical patterns that commonly match the meanings:

word meaning pattern example

High: (measurement from bottom to top) amount-adj/noun-adj: 2250mm high

High: ( used as a noun – ‘a high’) count noun/often noun-of-amount: a high of 394p

High: (high priority) adjective-on-noun: a high priority on technical training

Further, a combination of the COBUILD work and the BEC can introduce students to the typical patterning of business lexis using authentic examples:

word meaning pattern example

Sale: (sale of goods) singular noun: the sale of shares, sale of the goods

Sale: (sale in a shop) count noun: The January sales

Sale: (offered for people to buy) for-sale: approved for sale, available for sale

9.5.6 Students should know that words are not evenly distributed

Aspects of word distribution in the BEC have not been the main concern of this study, but what brief mention there has been has raised points that have pedagogical implications.

1. The genre-specificity of lexis: The distribution charts have clearly shown that some words tend to be used in some macro-generic areas rather than others, and that some very common words accumulate rapidly in certain lexical environments. The cluster Wall Street, for example, was only found in five macro-genres:

The chart above shows that most occurrences of Wall Street were found in US television programmes (ustv.txt), magazines & journals (mags&j~.txt), radio programmes (radio.txt), newspapers (wspa~1.txt) and meetings (meetings.txt). Another example, the word shall, was found to be heavily used in manuals and agreements:

The manuals where shall was commonly used displayed ‘agreement-like’ language with the terms of use being laid out for the users. Thus students’ awareness of language can be raised by pointing out these facts. Students needing to operate within a given genre, and perhaps more importantly, teachers needing to create suitable genre-specific materials, can see from lexical distribution which words are the most important for students to know and further study how they are actually used in practice.

2. Lexis for doing or talking about business: Pickett’s (1988) notion that business language can be divided into language for knowing about business and acting in business was found here to be correct, not just at the level of language routines as Pickett stressed, but also at the level of individual words. Materials writers need to be aware of the tendencies of words to be used in one domain or the other when creating teaching materials. However, much more work would need to be done on this before it could be put into practical use.

3. Lexis for speaking and writing: Likewise, the fact that some lexis tends to be used more in speech than in writing and vice versa, needs to be a conscious part of materials creation. This was found most commonly at the level of clusters where examples such as in order to were clearly on the written side and phrases such as a lot of and we need to fell on the spoken side. Again, more work needs to be done on this to see how a much larger range of words behave in the business environment.

9.5.7 There should be a greater focus on word clusters

The case for stressing the importance of both collocation and clusters for work in the classroom has been made by many writers (Lewis 1993, 1997, 2000, Williams 1998 and Hill 2000). Further suggestions for the use of clusters using the BEC are made here.

1. The generating power of small clusters should be exploited: Hill (2000) noted the importance of language chunks to learners, saying that their very predictability can be a powerful resource for the classroom. The BEC has generated those chunks that are found to be most commonly or most significantly used in the business environment. Students can be explicitly taught these clusters and also made aware of how smaller clusters are often part of much larger ones. Below, for example, the cluster the end of is shown enlarged into 6- and 7-word clusters:

Fig. 64 6-word clusters 7-word clusters

For examples of materials utilising clusters see Appendix 11 in Vol. II, p.902 (Exercise 5 – Exploiting business clusters).

2. The typical colligational patterning of clusters needs to be explicitly taught: The teaching of the smaller chunks mentioned above, and the way in which they fold out into larger chunks, has consequences for teaching. The typical grammatical setting in which the smaller clusters appear is made obvious by simply enlarging the base cluster. Thus, the cluster shown above, the end of , seems to fall into the pattern preposition-the end of-noun. Taking another cluster, one of the, it is commonly found in the pattern one of the-superlative, one of the-things or one of the-number. Taking the example of one the things, this can be expanded to form a larger grammatical patterning: one of the-things-that-(personal pronoun):

This can be the basis for a great deal of language work in the classroom, and by taking advantage of examples from the BEC, this rather ‘everyday’ cluster can be placed in a business environment. More importantly, in learning these clusters, students are avoiding making grammatical decisions of their own that could lead to error. Instead they are learning not just a fixed lexical form, but also a fixed grammatical chunk that can help them process the language more quickly and easily.

3. Clusters should be taught in terms of semantic prosody: Just as with individual words, clusters need to be presented to students in their most typical semantic environments. It was found with both 2- and 3-word clusters that they associated with clear semantic groups. Students could, therefore, be made aware that a very business-related cluster such as market leader is often found with lexis related to products/line of business:

and places:

and that even a seemingly more genre-independent cluster, such as a lot of, associates with business co-text when in the Business English environment, shown here firstly with companies:

and money:

9.5.8 Associate words should be more recognised

Scott’s (1997, 1999) notion of associate lexis deserves more attention both academically and in materials. To some extent, students have become used to working with collocations, and the introduction of the idea of associates to the classroom could provide a broadening of the lexical background to words. It can also be of great use for students who need to operate within a specific genre as the associate principle does not just show which words go together, it also shows words that are not found in each other’s company. This could be studied by use of Scott’s concept of clumps (1997), where associates are grouped by text to show their specific characteristics. Associates can be used in the following ways:

1. Associates can be used to help materials writing: For materials writers, associates can provide an important point of reference in terms of what co-text is presented with key words that they are writing about. It has been found in the PMC that the co-text used by writers thus far tends to skew the lexis away from that which is commonly found in actual business situations. Using the example of the word market, other key words that should surround it are shown below:

Thus, the words shown above here, e.g. business, companies and share, could take preference in materials when presenting the lexical environment of the word market.

2. Use of associates can add a further level of realism to materials: Materials should be seen to be collocationally, semantic prosodically, colligationally and now associatively correct. This might seem to place a heavy burden on the writers, but in fact, if a Business English corpus is used and authentic examples extracted, all this should happen automatically. It is only when writers start to invent examples and texts that inaccuracy or atypicality is more likely to occur.

3. Associates can help set the lexical syllabus: The lexis surrounding key key-words can be used to direct students into more effective learning. When lower ability-level students learn business lexis they not only need to do so relatively quickly, they need to be sure that what they learn is relevant to the business environment. The use of associates allows them to see the lexical landscape in which a key key-word is set, and teachers can use this fact to decide what needs to be taught next. If the lexis concerning customers is being taught, for example, alongside the collocation, the semantic prosody and the colligational patterning, the teacher can introduce other words that are found to be satellite to the word customer:

These associates can then be chosen for more in-depth teaching in their own right.

9.5.9 Pedagogical issues: a conclusion

To summarise this part of the chapter it has been found that work on the BEC has both confirmed previous studies in the area and also raised some new points. Ten points can be made in conclusion.

1. Intuition has been shown to be flawed with regard to Business English materials and it is recommended that materials from now on use a corpus of Business English as their starting point.

2. The key words found in the BEC can form the basis for new materials. They provide a store of words that are a statistically significant part of Business English and offer opportunities for both teachers and students alike to explore the lexis of business with a higher degree of certainty that the words they are studying are relevant. However, the key words must be seen only as a point of departure rather than an end in themselves. Any attempt to artificially limit the lexical input of Business English would be ultimately self-defeating. No corpus is perfect, and lexis that has been missed due to possible deficiencies in this corpus should not be dismissed out of hand.

3. The fact that Business English is to a large extent made up of a limited number of semantic groups of lexis needs to be specifically taught to students and made an overt part of materials. These semantic groups should be a central focus of materials but not the only focus. Again, Business English is larger than these groups and this, too, should be taught to students.

4. There needs to be a higher priority given in materials to words that are in the background of business, words that are not pure business but words that students need in order to take part in business life, for example, plan, forecast, control and issue. Materials so far have featured them, but focus on these words needs to be made an important part of Business English teaching materials. Likewise, there needs to be a greater stress in the materials on business-related verbs as opposed to noun/verbs. Use of the BEC can help this process.

5. The results of the analysis by semantic prosody on Business English lexis need to be utilised in materials design in the future so that words would be placed in an accurate lexical setting. This has been done to a reasonable extent in previous materials, but more accuracy, increased lexical richness and an explicit reference to semantic prosodic groups needs to be made in materials of the future. Additionally, more precision is needed in materials with regard to colligational patterning. Typical patterns of words need to be taught to students and these should be gained from empirical data.

6. The concentration in the published materials on travel, food, entertaining, and personal contact needs to be broadened out to cover other areas of business life and, for example, the focus on politeness could be offset by a focus on speaking and writing positively.

7. The relationship between grammar and meaning should be made more explicit to students – the fact that meanings have typical grammatical patterning, e.g. the verb-on pattern with improve on (improve on a previous achievement) – and specialist business meanings of everyday words could be more stressed than has been so far (e.g. partner, confirm).

8. Materials could take into account the distribution of words across and beyond genre and ensure that they are placed in environments that are typical of them – e.g. in order to was found to be more common in doing business and one of the more common in talking about business.

9. Clusters could be exploited more in materials and their power to generate longer chunks of language made clear to students – e.g. the end of /at the end of/ at the end of the day.

10. Associates can be used to add extra substance to Business English materials and can facilitate, to some extent, the setting of the syllabus – thus when learning about the word customer, the associates business and products are found to co-occur and could be candidates for words to teach next.

9.6 Incidental findings

The thesis has largely been limited to analysis of key words, which has meant that many other lines of investigation have been limited or excluded. During the course of data collection and writing, several issues were noticed that, although related to the main study, time and space have forced to the sidelines. These will be briefly mentioned here and could possibly form part of future work done on the BEC.

Gender bias: The BEC has a stronger male bias than the PMC, where lexis related to women is more evenly represented in the texts. It was found that in the PMC, Mrs, ma’am and madam were all positive key words, whereas female-related lexis was not key in the BEC – indeed Mrs, for example, was a negative key word in the BEC. The frequency table below, taken from the BEC, indicates the male bias in the BEC with the relative frequencies of references to male and female lexis:

TABLE LXXXII: EXAMPLES OF THE RELATIVE FREQUENCY OF MALE-FEMALE LEXIS IN THE BEC

Male	Number of instances	Female	Number of instances
man	130	woman	36
Mr	1,020	Mrs	33
his	912	her	259
son	16	daughter	11
gentleman	12	lady	12
gentlemen	5	ladies	27

It can be seen that the ‘male’ lexis outnumbers the ‘female’, for example, 130 mentions of man and only 36 of woman. The only female word shown above found in the BEC that was more frequent than the corresponding male one was ladies.

Swearing: No swearing was found in the PMC. A very limited amount of mild swearing was found in the BEC, and whilst stronger swear words were absent from the BEC, they could be found in the BNC. Thus it appears there is a sliding scale: published Business English materials attempt to be polite, real-life business allows some mild swearing amongst people of previous acquaintance, and finally, everyday life, where swearing is more commonplace. This is shown in the table below, followed by examples from the three corpora:

TABLE LXXXIII: THE SLIDING SCALE OF USE OF SWEARING:

PMC ->BEC ->BNC

PMC – materials >>	BEC – real Business >> English	BNC – everyday English
no swearing >>	mild swearing >>	mild and strong swearing

Polite language in the PMC – there is not even one instance of shut up or be quiet, for example:

Mild bad language in the BEC:

Strong swearing in the BNC:

This sliding scale of language use also ties in with the materials’ predilection for politeness, and further demarcates the world of business by showing that mild swearing is acceptable in certain circumstances, for example, bloody, shit, sod, but the strong swear words of everyday life are seemingly not acceptable in business.

Reported speech: Work by Yule, Mathis & Hopkins (1992) found that when reporting spoken speech, people ‘construct dialogue for the participants. The result is often a presentation, in direct speech form, of a recreated interaction in which thoughts and attitudes, as well as speech, are reported’ (Yule, Mathis & Hopkins 1992: 248). When transcribing the tapes for the BEC this feature of spoken language stood out. Reported speech in informal meetings rarely took the form taught in text books (changes in pronouns and temporal adverbs, backshifting in tenses, and the position of the reporting verb) e.g. She said that she would meet me at 10 o’clock. Instead, the person being talked about was almost always directly quoted word-for-word, with an introductory phrase preceding or following it, such as She said:

This is an area that could be looked into in more detail in the future and certainly an area that needs to be given more attention in Business English teaching materials.

Hedging and discourse patterns: The lack of reality in business materials with regard to discourse patterns and lack of hedging devices has been noted in the literature, most recently by Kirk & Tamkin (2000) in their work on business meetings. Two examples from the different corpora serve to confirm the differences. The first example is a telephone conversation taken from Business Opportunities (Hollett 1994):

<A> Alan Wilson.

Good afternoon. This is Cristina Garcia. I’m phoning about our appointment on the 16th. I’m sorry, but I can’t make it.

<A> Oh dear.

Can you manage the 18th instead?

<A> I’m afraid I’m tied up then.

The only other time I’m free is Thursday morning. That’s the 19th.

<A> Well, I’ve got another appointment but I can cancel it. Would 10.30 be

convenient?

That’s fine. I’m sorry to be a nuisance.

<A> It’s quite all right.

I’ll see you on Thursday at 10.30, then. Thank you, Mr Wilson.

<A> Good bye.

The second example is taken from a real life phone call to a sports goods company:

<A> Good morning, Companyname.

Good morning. My name is Personname from Companyname, in Placename.

<A> Hello there.

How are you, all right ?

<A> Yes, I’m fine thank you.

Well we, I had an order come recently that it was, it was er…all ladies er…size 36 vests. But there was one missing.

<A> Right.

I just, I just wondered how soon that one would be down ?

<A> Erm .. well I would say, you know, that we’ve had to re-order stock so ….

I see, yeah….

<A> Erm …. I would say they’d probably be about another, you know, ten days at the moment – I would say.

Oh fair enough.

<A> Is that all right?

Yeah, but obviously if it says ‘balance to follow’ that’s it and if it’s just the one …

<A> Yeah.

…..because everything else came down ….

<A> Right

….. because there was only the one shirt missing.

<A> Yeah… so it’s just obviously if we’ve run out of that particular size….

Yeah…

<A>… you know we just have to then re-order I’m afraid.

All right, yeah .. funnily enough it was all 36s I was getting – for some reason I seem to have a rush on it.

<A> Did you? Yeah.

Okey doke.

<A> OK so you still want that one there’s… yeah.

Yeah.

<A> You’re not, you’re not adding at the moment ?

No, no that’s just, I just wondered because I’m going to the club tonight and I know …

<A> Yeah, sure ..

… people always ask then ‘How are we because so and so have got theirs….’

<A> Yeah, yeah.

‘and mine’s not come’. That sort of business, you know. The usual story.

<A> Yeah. Just tell them it won’t be, it won’t be too long, just be patient.

That’s lovely.

<A> OK ?

Thanks very much.

<A> Sorry for the delay anyway.

That’s OK.

<A> OK. Bye.

Bye.

In the first conversation, taken from the book, the dialogue is presented as a linear event – one person speaks, the other listens, and there is no overlapping. The second phone conversation shows constant overlapping and is a decidedly non-linear event. Additionally, the implicit/explicit language argument noted in Chapter 3, for example by Boxer & Pickering (1995), can be seen very clearly here. In the book phone call, direct questions are answered by direct questions, e.g. Can you manage the 18th instead? – I’m afraid I’m tied up then. In the real phone call, answers are less direct, e.g. I just wondered how soon that one would be down? – Erm, well I would say, you know, that we’ve had to re-order stock so. The language is full of hedging and hesitation devices: erm, you know, it is not overly polite for example, the seller says just be patient, and the seller often checks understanding with phrases such as OK, right and all right. Several false starts and changes in mid-sentence can also be found, e.g. Well we, I had an order.

Obviously some simplification in materials is justified, especially for lower level students, but in general, the phone conversations found in the materials do nothing to prepare students for phone calls that they are actually going to have. Students ought to be prepared for the chaos of telephone conversations. When transcribing the tapes it was notable how often the two people were actually talking at the same time, yet still somehow being able to listen and take in what the other person was saying. Materials need to take this into account and slowly get students used to it. Kirk & Tamkin suggest that ‘it is essential that we equip our students with the strategies they need for this kaleidoscope of language. Great emphasis should be laid on helping the student to formulate questions asking for clarification, reiteration and even simplification of statements or questions’ (Kirk & Tamkin 2000:1).

Vague language: In a related issue, Joanna Channell’s (1994) concept of vague language can be found in abundance in the BEC. Channell goes to some length to show that the very concept of vague language is itself vague (Channell 1994:18), but it can be simplified as being language that is used in the place of more specific language for a variety of reasons. Some brief examples are given below:

Here again is an area that materials writers should take more note of and prepare students for the complexities of everyday language.

Public face – internal face: The language used inside a company amongst specialists, and the language used outside, to the public, is different. This idea is not new and was discussed at length in Chapter 3, notably in reference to the work of Pickett. Work in gathering the tapes for the spoken part of the BEC, however, showed that there is also a parallel phenomenon in terms of attitude and language. In private, business people talk about what actually happens in business, in public, the company face is presented. Compare the following extracts:

I don’t say you have to lie, you just embroider the truth.

In theory we should stick to ISO 9000, in practice we go either side of the line.

What do we do? We screw ‘em don’t we? We all do it.

I’ll come over there and beat you up.

All these were heard either on or off tape during the data gathering process. Compare these to a sample of business people talking about their businesses found in published materials (Business Opportunities – Hollett 1994):

I wanted a business I could fall in love with.

I picked the best one.

Microsoft encourages individuality.

They may get it wrong several times before they get it right but they never give up.

The materials thus present the picture of companies that they would like to present to the outside world. The emphasis on positive language in the BEC has shown that this language does form a major part of Business English. However, students could also be prepared for other, more honest situations.

Socialising in meetings: Socialising in meetings is often presented in materials as a pre-meeting chat before somebody says ‘let’s get down to business’. This can be seen clearly in the extract below taken from New International Business English (Jones & Alexander 1996):

<A> Ah, good morning, Ms Ross, do come in.

Hello, Mr Fisher. Nice to see you.

<A> Nice to see you – face-to-face instead of on the phone, what? How are you?

Fine, thanks, very well.

<A> Oh, do sit down. Would you like some coffee?

Oh, yes, please – black.

<A> Mmm. Here you are.

Thanks. Well, how’s it all going?

<A> Oh, not too bad, we’re just about to open a branch in New Zealand.

Oh, will you be going there on your travels?

<A> Oh…I’m hoping to – if I can justify it to the marketing director! How’s your little boy, has he started school yet?

Oh, yes, he’s in the second year now.

<A> What, already? Doesn’t time fly! Is he enjoying it?

Very much, it’s much more fun than being at home!

<A> Haha. Well, I suppose we’d better make a start. Shall we get down to business?

Right. First of all can I confirm the time and date of the presentation? It’s Saturday 24 October in the morning.

<A> What time exactly?

Small talk is followed by the main purpose of the meeting. The real life meetings showed that although social chat did occur at the beginning of a meeting, it also came up repeatedly and formed an integral part of the actual body of the meeting. It was used to relate aspects under discussion with more social aspects. In the extract taken from a meeting below (Meeting 3 in the BEC), the social aside is integrated into the main purpose of the discussion:

<A> Getting rid of all this hassle is causing us quite a lot…

I know, I know. I mean, the thing is …

<A> Until we get it in place.

I mean, once it’s in place hopefully it should, you know, be OK.

<A> Interestingly, I was talking to Personname the other day, or more that she was talking at me, because she’s just lost her job. She’s been er, she’s worked for National Express coaches, and has been for a long time – running a large part of Scotland. Er, and it was obvious to her it was going to be taken over and the job would disappear, so she moved to something else and then that collapsed as well, so she’s on a brief period of unemployment. At least she hopes it is, but she’s it’s interesting that what she’s looking for, which is just jargon I wouldn’t have used, ‘I’m not looking for anything where I have to get new business all the time, my, my strengths lie in account management.’ And basically all she’s been doing over the years is just keeping the customers happy and keeping her boss happy that the customers are happy kind of thing. Account management. Erm, and that’s basically what we’re doing isn’t it? We are trying to get hold of this we’re trying to manage it over one or two meetings with you and me, thereby reducing lots and lots and lots of trivial work.

The integration of socialising and the main purpose of the meeting is found again in another meeting (Meeting 2 in the BEC), this time placed at the beginning. There is no break in conversation of a ‘get down to business’ type – the business simply flows from the socialising:

<A> So, how’ve we been, all right?

Erm, it’s been a crazy four weeks,

<A> A crazy four weeks.

Yeah. We’ve had this three weeks maybe, and erm and it’s just now we’re starting to get somewhere…….

<A> Because you’re working forward, you mean, or or just …..

Yeah, we’re working forward ….

<A>…. or you’re not talking, dealing with current sales needs are you ?

Most of the current stuff we’re, we’re the, the emphasis is, is turning slowly erm, in the past the idea was we would never know what we would sell, so we stocked absolutely everything for everybody. And that was then left for a line of clearance and that creates problems because after a while it’s inbred to wait for the clearance. And now we are buying more and more to order which means there’s less and less stock lines so it becomes a different type of problem. Er more, it’s a certain line, where we’ve got customers screaming for the product and there’s no credit because it’s all sold.

<A> Right.

There is also a marked contrast in formality between the meeting extract from the book (quite formal: good morning Ms Ross, do come in) and those taken from real life (informal: So, how’ve we been, all right?), where in both situations the people have had previous contact before, albeit in the book, only by phone.

Formality of business letters: Louhiala-Salminen (1996) noted the changes in styles of business correspondence over the last decades and pointed to the decrease in formality in business writing. Based on the data gathered for the BEC, this informality was found in, for example, emails, but a cursory look at the business letters still reveals a good deal of formal language. Some examples are shown below:

Dear Sir

We are writing to formally set out our responsibilities as auditors and our understanding of the additional services you require us to perform.

Dear Mrs Personname

I do trust that you are receiving a consistently high standard of cleaning from the Teamleaders who are looking after your premises. However should you ever have any cause for complaint please do not hesitate to mention this to the Teamleader or alternatively contact our office. This enables us to monitor the situation and ensure that the high standard is maintained.

Dear Mr Personname,

Further to your telephone conversation with my colleague Personname, please find enclosed information on the activities of ….

This suggests that more traditional phrases are still in use, and that teaching materials should still reflect this reality.

9.7 Critique and future applications of the corpora

One of the key aims of this whole project was to create corpora that could be used by a larger number of researchers. To this end, both the BEC and the PMC will be presented to the University of Manchester for research use on the completion of this project. Possible areas for on-going study using the corpora will therefore be suggested at the end of this section. The first part, however, will focus on possible criticisms of the research.

9.7.1 Critique

a) Ideal vs Actual BEC content: It has been seen that the planned BEC corpus differs from the corpus finally collected, and several compromises had to be made along the way. In the spoken-doing part of the corpus there is not as much business-customer dialogue as would have been liked, with the majority of the text coming from internal meetings. Companies seemed to have fewer qualms about giving away company-customer written correspondence, of which there is a lot in the BEC, but had serious misgivings about allowing direct access to spoken events. The tapes recorded for the BEC do have direct customer contact in the form of a rep/customer meeting and tens of telephone calls. There are also four company/distributor negotiations, which display inter-company discussions. The rest of the spoken-doing part of the corpus is made up of internal meetings, and the lack of customer involvement is compensated for by the variety of different meetings, including sales, finance, product and technical meetings. It should be remembered, also, that a large percentage of business people have little or no direct spoken contact with customers/clients, and in this respect the BEC is not as lacking as might be thought.

One definite lack in the BEC is that of overt socialising situations and it proved impossible to gain access to business lunches, for example. Socialising, however, as mentioned above, can be found in the corpus, but is integrated into meetings, and takes place largely between people who have known each other for while. Both Williams (1988) and Charles (1996) have studied the effect the interlocutors’ acquaintance has on the language used. Thus, this aspect of business is given little focus in this study.

A further criticism is that not all macro-genres achieved the adopted minimum sample size of 20,000 words. In all, four macro-genres fell short of the target – negotiating 16,000, training session 17,000, memos 12,000 and quotations 8,000 words. Speeches and presentations also fell short, but only by just under 1,000 words. Despite not reaching the specified 20,000 words, these texts were included as they added to the balance of their corpus. Further, no separate statements were made about these individual genre due to the small sample size, and so their presence in the corpus can only be seen to have a positive effect. For a fuller discussion on ideal vs actual BEC content, please see Appendix 12 in Vol. II.

b) Problems of text placement: There was some discussion in Chapter 3 on the difference in naming types of documents between business people and researchers. Barbara et al. noted that ‘the same label is being used to refer to different document types, and conversely, that different labels may be referring to the same document type’ (Barbara et al. 1996:69). Similar problems were faced with this project, notably with electronic messages. An email is an email, a fax is a fax, but it needed to remembered that these are both simply methods of information transmittal, and not fixed document types in themselves. Thus an email or fax may contain a report, a quotation or a simple three line message. It was attempted to place all similar document types together, no matter what their mode of transmittal. More problematic were other text types, and on several occasions the business people supplying the documents had to be asked exactly how they would categorise them. Other documents displayed features of two document types and so it was difficult to place them in the corpus. For example, some manuals, whilst giving instructions, also made stipulations of usage that placed them lexically close to contracts. In all cases a decision had to be made on where to place texts, meaning that future researchers may wish to re-assign certain texts to different parts of the corpus.

c) Intuition, lexis and semantic groupings: Computerised generation of statistics can only go so far in a study of this nature. Statistics need to be interpreted and the use of a researcher’s intuition is necessary. Much of this thesis rests on the semantic groupings made of the key words and this was achieved by use of intuition. It is therefore potentially flawed, and there were problems of lexical assignation especially between some categories, for example, between lexis to denote events and lexis to denote activities. There is no doubt that the semantic groups are supported by relevant lexis, but some lexis may have been placed in more than one category. Likewise, it may have been possible to create further semantic groupings.

d) Manual analysis of semantic prosody and colligation: A similar problem was encountered with the analysis of semantic prosody. The concordance lines for each of the 50 words studied were printed out. These could amount to over 100 pages of concordance lines for just one word, as in the case of business, and were very often between 20 and 30 pages in length, with 50 concordance lines being found on each page. This posed two problems. Firstly, the finding of semantic prosodies had to be done completely manually, and was dependent on simply what the eye could see. This leaves open the possibility of alternative prosodies being found by other researchers who may pick up on aspects not noticed here. Secondly, the large amount of examples that had to be counted by hand also means that slight errors in the number and percentage value of the prosodies given, despite being checked three times, could occur. A related problem was also found in analysing colligational patterning, where the tagged corpus was only of marginal help. The counting and noticing of grammatical features was almost totally manual, meaning that similar errors as noted above with regard to semantic prosody are possible. Finally, division of COBUILD meaning categories was also done manually, and placing examples of words into meaning categories was at times difficult.

e) Breadth of the study: This work has cut across many fields of study: register, discourse and genre analysis, corpus linguistics, lexical, collocational and multi-word item studies. Each of these areas has potential for a great deal of work. This thesis, in some ways, has skimmed the surface of these areas to achieve its aims. This can possibly be seen as a minus – no one area has been covered in great depth. On the other hand, it can also be seen as a plus – many differing viewpoints have been included that affect the final outcome.

f) BEC – PMC differences: Discussion in the last chapter with regard to the work of Biber (1988) has shown differences between the BEC and the PMC in terms of grammatical and textual characteristics. The BEC tended to the informational end of the informational vs involved production scale, whilst the PMC, displaying a high number of personal verbs and first and second person pronouns, tended towards the involved production end of the same scale. Although Biber states that this scale ‘is not in terms of oral and literate discourse’ (1988:108), he does suggest that it ‘marks the extent to which they are oral or literate in terms of their production characteristics and primary communicative purposes’ (1988:108). It could be suggested, therefore, that the lexical differences found between the BEC and the PMC primarily result from more spoken language being found in the PMC than in the BEC. It is certainly true that the PMC contains a good deal of spoken language, for example, whilst only three books in the PMC are devoted specifically to writing (Handbook of Commercial Correspondence, Company to Company and Written English for Business II), totalling 57,000 words, ten books are specifically devoted to speaking skills – totalling 83,000 words. The remaining twenty books are divided between spoken and written language. Thus, it is possible that the PMC is biased towards spoken language, whilst the BEC consists of 44% spoken texts (447,318 words) and 56% written (575,703 words).

It should be noted, however, that similar differences between the PMC and the BEC were found irrespective of the reference corpus being used. Thus, the over-positive, over-polite, over-personal key words found in the PMC using the BEC as a reference point were also found when using the BNC as reference – though not to the same extent. This is significant, as the BNC has a more or less equal split between spoken and written texts and so matches the make-up of the PMC more closely. Also, the differences between the two corpora were not limited by characteristics belonging to Biber’s factors – there was an essential semantic difference found between the PMC and the BEC that transcended Biber’s work, for example, different types of people and a stress on travel and entertaining. Furthermore, as pointed out earlier in this chapter, the findings of this research closely match earlier research on published EFL materials. The essential point to be made therefore, is that the PMC focuses on language that is concerned with person to person interaction, seemingly in both written and spoken form, whereas the BEC, and, consequently business lexis, shows a much greater focus on informational aspects of communication.

9.7.2 Further applications

In this thesis only very limited use has been made of the two corpora, and several of the aspects discussed could be examined in much greater detail in the future. Below is a list of suggestions of work that could be done to both continue on from this thesis and also expand research into other areas.

1. The work done on semantic prosody could be continued to encompass all the key words found, in addition to the 50 words chosen in this thesis.

2. Likewise, the colligational patterning of all the key words could be explored.

3. A full analysis of semi-business lexis could be carried out by continuing analysis of words in relation to meanings found in COBUILD. In this way the business-specific meanings of general lexis could be comprehensively documented for inclusion in future teaching materials.

4. Much more work could be done on the clusters and chunks of language found in the BEC along the lines of Williams (1998), where she classified the clusters and adopted them for use in the classroom.

5. Further work also needs to be done on associates, and on taking advantage of the lexical insights they give. For example, they could be used in helping to lexically define genre as they statistically mark out which words go together and which words do not.

6. A large amount of work could be done on comparing native speaker business language to non-native speaker by using the BEC as a reference corpus. Thus, for example, non-native speaker meetings could be compared to meetings recorded for the BEC. Comparison could be on many levels – register, discourse and genre to mention but a few.

7. The macro-genres found in the BEC could be further broken down into individual genres as used as the basis for further study.

8. The discourse patterns found in the spoken sections could be the basis of further studies, for example, centring around the use of hedging and hesitation devices.

9. Sexism in business language could be further explored.

10. Finally, the BEC itself stands as a broad bank of information on the lexis, grammar and semantics of Business English. It is hoped that this bank can be used by researchers to gain further insights into an area that has so far largely been covered in shadow.

[196] Associates could not be computed in a similar manner to words in the BEC, as the PMC was categorised by book and not macro-genre. Associates, though computed, were not included in the analysis of the PMC.

[197] It is important to note again here that the BNC is being used purely as a statistical reference point. Analysis is not concerned with the BNC, but rather the words in the PMC that occur statistically more frequently than in the general English corpus of the BNC. It is these words that occur unusually frequently in the PMC, when compared to the BNC, that are ‘special’ to the PMC (published materials) and so warrant closer analysis. Discussion in the following sections compares these PMC key words to the corresponding key words of the BEC, which were also computed using the BNC as a statistical reference point.

[198] First and second person pronouns, indeed, are part of Biber’s (1988) Factor 1 noted earlier, denoting a focus in the language on interpersonal matters.

[199] Probably key for its US spelling rather than intrinsic value.

[200] The word termination also appears here, but has not been included as being purely negative, though, of course it can have negative connotations.

[201] The words computer, software, telecommunications and technology, were, however, key.

[202] Both of these were, however, very low in frequency, so safe conclusions cannot be drawn from them.

[203] In the analysis of the PMC, negative key words – that would show the words under-represented in the PMC – have not been a major part of the analysis. They are, however, referred to in Section 9.4.2.4 later in this chapter.

[204] In the same way that Ljung (1990) used under- and over-representation in his difference coefficient.

[205] This refers here to the lemma – so all inflections of the word are included, e.g. customer, customers. The

figures in the semantic prosody table refer to just the word customer only and so are less in number.

[206] It has already been noted in the analysis of key nouns that the PMC over-represents certain jobs, manager being one of them.

[207] Right collocating groups only.

[208] For more on this see Section 9.4.5.2 on 3-word clusters that follows the grammar/meaning discussion.

[209] This will be also dealt with in the next section.

[210] This sense was contrasted in COBUILD with specific markets (Sense 2), e.g. the market for luxury holidays.

[211] One further difference between the PMC and the BEC that should be made here is between positive and polite language. The BEC was found to stress positive language, whilst the PMC was found to stress polite language and more equally represent positive and negative language.

[212] This incident happened to the author in spring 2000.

[213] Further, the fact that these words are computer-based facilitates exploitation of the differing linguistic characteristics they possess.