Structure of the WSC (1988-1994)

1.  Categories of the WSC

The WSC comprises different proportions of formal, semi-formal and informal speech. The extracts are divided into 15 categories and these categories cover a range of contexts in which each type of speech is found. In table 8.1, WSC Categories and Word Targets, the categories are grouped in terms of whether they are monologues or dialogues, public or private, scripted or unscripted. The codes assigned to the categories are also provided, along with the word targets for each category.

The formal speech section of the WSC involves all the monologue categories and the DGUs (Parliamentary debate). The semi-formal section is comprised of the interview categories, both public and private: oral history (DPH), social dialect (DPP) and broadcast interviews (DGI). The remaining dialogue categories comprise the informal speech section, with 50% of the overall corpus being comprised of private face-to-face conversations (DPC).

Table 8.1: WSC CATEGORIES AND WORD TARGETS

Category	Text Category	Code	Word Target
Monologue	Broadcast news	MSN	24,000
Public scripted, broadcast	Broadcast monologue	MST	10,000
	Broadcast weather	MSW	2,000
Monologue	Sports commentary	MUC	20,000
Public unscripted	Judge's summation	MUJ	4,000
	Lecture	MUL	28,000
	Teacher monologue	MUS	12,000
Dialogue	Conversation	DPC	500,000
Private	Telephone conversation	DPF	70,000
	Oral history interview	DPH	20,000
	Social dialect interview	DPP	30,000
Dialogue	Radio talkback	DGB	80,000
Public	Broadcast interview	DGI	80,000
	Parliamentary debate	DGU	20,000
	Transactions and Meetings	DGZ	100,000
TOTAL			1,000,000

Table 8.2, WSC Categories – Targets and Final Figures, shows the number of words actually collected for each category, as well as the number of extracts. The WSC consists of extracts of approximately 2,000 words (as used by the Brown, LOB and ICE corpora). Exceptions to this target are made when the entire speech event is less than 2,000 words (e.g. weather reports, shop transactions and news bulletins).

Table 8.2: WSC CATEGORIES – TARGETS AND FINAL FIGURES

Code	Text Category	Number of Extracts	Word Target	Words Transcribed
MSN	Broadcast news	36	24,000	28,929
MST	Broadcast monologue	5	10,000	11,205
MSW	Broadcast weather	12	2,000	3,641
MUC	Sports commentary	10	20,000	26,010
MUJ	Judge's summation	2	4,000	4,489
MUL	Lecture	14	28,000	30,406
MUS	Teacher monologue	8	12,000	12,496
DPC	Conversation	226	500,000	500,363
DPF	Telephone conversation	46	70,000	70,156
DPH	Oral history interview	10	20,000	21,972
DPP	Social dialect interview	11	30,000	31,058
DGB	Radio talkback	37	80,000	84,321
DGI	Broadcast interview	40	80,000	96,775
DGU	Parliamentary debate	14	20,000	22,446
DGZ	Transactions and Meetings	80	100,000	102,332
	TOTAL	551	1,000,000	1,046,599

The word counts for some of the categories include words from individuals whom it was not possible to contact for permission or background information sheets (see section 11.1.1, Who counts as a New Zealander?). The MSN Broadcast news category includes 709 words from such people (2% of words in this category). The DGB Radio talkback includes 49,016 words from such people (58% of words in this category). In all other categories the number of words contributed by such people is negligible.

More information on the different categories in the WSC is provided in section 15, Texts, along with information on each extract included.

The word counts quoted in this guide are based on DOS word counts produced from the original wordperfect files. These files have been converted and reformatted for the release version of the corpus. Word counts in the release version, therefore, may differ slightly.

2.  WSC Gender, Ethnicity and Age Breakdowns

The following tables give the final figures for the number of words in each category with a breakdown by gender and by the two main ethnic groups represented - Pakeha and Maori. The age breakdown for the overall corpus is shown in figure 8.1, Age Composition of WSC. In this section, the figures for several of the categories – especially MSN and DGB - do not match the figures in table 8.2, WSC Categories – Targets and Final Figures, because we do not have demographic information for all participants.

Table 8.3: WSC Composition by gender

Code	Text Category	Overall Words	Number from Females	Number from Males
MSN	Broadcast news	28,166	10,114	18,052
MST	Broadcast monologue	11,205	4,453	6,752
MSW	Broadcast weather	3,641	388	3,253
MUC	Sports commentary	26,010	0	26,010
MUJ	Judge's summation	4,489	0	4,489
MUL	Lecture	30,406	11,151	19,255
MUS	Teacher monologue	12,493	9,479	3,014
DPC	Conversation	500,332	301,521	198,811
DPF	Telephone conversation	70,156	50,554	19,602
DPH	Oral history interview	21,972	12,760	9,212
DPP	Social dialect interview	31,058	14,083	16,975
DGB	Radio talkback	35,304	6,554	28,750
DGI	Broadcast interview	96,696	36,043	60,653
DGU	Parliamentary debate	22,446	6,349	16,097
DGZ	Transactions and Meetings	102,122	52,826	49,296
TOTAL		996,496	516,275	480,221
			52%	48%

The WSC data was collected between 1988 and 1994. The New Zealand overall population figures from the 1991 Census show that 51% of the population was female and 49% male. (Census figures are taken from New Zealand Official Yearbook, 95^th edition, Department of Statistics 1992.)

Table 8.4: WSC Composition by ethnicity

Code	Text Category	Overall Words	Number from Pakeha	Number from Maori
MSN	Broadcast news	28,166	20,300	7,204
MST	Broadcast monologue	11,205	11,205	0
MSW	Broadcast weather	3,641	3,641	0
MUC	Sports commentary	26,010	24,732	0
MUJ	Judge's summation	4,489	4,489	0
MUL	Lecture	30,406	26,315	4,091
MUS	Teacher monologue	12,493	10,360	0
DPC	Conversation	500,332	366,047	92,154
DPF	Telephone conversation	70,156	62,985	1,689
DPH	Oral history interview	21,972	21,972	0
DPP	Social dialect interview	31,058	706	30,352
DGB	Radio talkback	35,304	31,226	1,765
DGI	Broadcast interview	96,696	56,735	39,466
DGU	Parliamentary debate	22,446	22,257	189
DGZ	Transactions and Meetings	102,122	92,772	3,771
TOTAL		996,496	755,742	180,681
			76%	18%

Ethnicity figures from the 1991 New Zealand Census show Pakeha constitute 73.8% of the population and Maori 12.9%. In collecting the WSC an effort was made to ensure that a reasonable proportion of the data was collected from Maori (see section 11.12, Ethnic and Gender Representation).

Figure 8.1: Age Composition of WSC (Number of words by age group)

Figure 8.1, Age Composition of WSC, shows the number of words contributed to the WSC by each age group. In figure 8.2, Age Comparison for WSC and New Zealand Population, an age comparison between the WSC and Overall New Zealand population figures is provided (estimated for 1990 from figures in the New Zealand Official Yearbook, 95^th edition, Department of Statistics 1992). WSC figures show the percentage of words contributed to the corpus by each age group, while the overall population figures show the percentage of the adult population in each age group.

Figure 8.2: Age Comparison for WSC and New Zealand Population

3.  WSC and ICE-NZ OVERLAP

As mentioned earlier, WSC and the spoken component of ICE-NZ share 9 categories. The following table lists the categories which are shared, the number of words collected for each corpus and the actual number of words which are shared. WSC extracts which are included in both corpora are identified in section 15, Texts.

Table 8.5: WSC and ICE-NZ OVERLAP

Code	Text Category	Words WSC	Words ICE-NZ	Actual Overlap
MSN	Broadcast news	28,929	40,396	26,401
MST	Broadcast monologue	11,205	45,276	11,205
MUC	Sports commentary*	26,010	52,007	26,010
MUJ	Judge's summation	4,489	22,473	4,489
DPC	Conversation	500,363	206,624	203,864
DPF	Telephone conversation	70,156	22,688	22,688
DGI	Broadcast interview	96,775	21,810	0
DGU	Parliamentary debate	22,446	22,446	22,446
DGZ	Transactions and Meetings	102,332	22,145	22,145
TOTAL				339,248

*ICE-NZ's commentary section is not limited to sports commentary.

The spoken component of ICE-NZ is still being finalised, so these figures are not final.