Corpus-based approaches to specialised vocabulary in the trades
Vocabulary in the trades
Several corpora were gathered for the LATTE project. The purpose of the written professional corpus was to find out more about the nature and frequency of lexical items in the texts which learners were exposed to in their studies in the trades. The written corpus was gathered by interviewing tutors to find out more about the texts they used in class and where possible, gathering these texts into a corpus for analysis. Table 8.1 shows the breakdown of the running words in the written professional corpus. Note that only texts which were actually used in the trades’ courses were included in the corpus, which means that the overall number of running words is fairly low at 1,641,000. In terms of balance between the overarching trades, Construction and Engineering, there are over 90,000 more running words in Construction than Engineering. Carpentry and Fabrication have the lowest number of running words.
What kinds of vocabulary might be in this professional corpus? To give a sense of the specialised lexis, Figures 8.2 and 8.3 provide examples of texts from two trades: Carpentry and Automotive Engineering. Figure 8.2 shows a short section of the Unit Standard text (121 words or tokens) about safe work practices on a construction site for New Zealand Carpentry students (New Zealand Qualifications Authority, 2017). This text is an example of a text from Carpentry for assessment purposes.
This sample of text is probably quite understandable for a general audience because its subject is health and safety. An analysis of this text found that it contains a range of lexical items, including around almost 92% from the first 4,000 words of Nation’s (2006, 2013) BNC/COCA 25,000 lists. Only four items, accordance, respirator, UV and extinguisher, occur outside the first 4,000 words of the BNC lists. Note the repetition of ’employer’s safety procedures’ in the text as an example of a multi-word unit.
Table 8.1 The written corpus of the LATTE project
Figure 8.2 A section from Unit Standard 13036, carry out safe working practices on construction sites
Figure 8.3 A sample of text on diesel from a textbook in Automotive Engineering (Weltec, 2016)
In contrast to Figure 8.2, Figure 8.3 contains 137 tokens from a set of materials on Diesel in Automotive Engineering. Like the example in Figure 8.2, just over 90% of the tokens in this text are in the first 4,000 word lists of the BNC, including mounted, port, pump, pumps, reduces,seal, transfer and trapped from BNC-COCA-2,000; consists, distributor fuel, injection, input, squeezed and volume from BNC-COCA-3,000; and offset, outlet, rotates, shaft and slots from BNC-COCA-4,000. There are two items which occur outside the first 4,000 lists of the BNC/COCA, and they are rotor in BNC-COCA-7,000 and vane/vanes in BNC-COCA-11,000. Note the multi-word units in the Automotive text, including distributor type injection.
Table 8.2 Examples of high frequency specialised vocabulary in Plumbing, Fabrication and Carpentry
Table 8.2 shows some more examples of vocabulary in Plumbing, Fabrication and Carpentry. When looking at this table, the first point to consider is whether people outside these fields of expertise would recognise these words, or use them in their daily communication. A word like sarking, in the third column of Table 8.2 or seaming in the first column, are unlikely to be well known outside of the trade. These words would clearly be in Step 3 or 4 of Chung and Nation’s (2004) scale of technical vocabulary (see Chapter 2). This point is important for learners and teachers who are focused on early learning in the trades because it illustrates the specialised nature of the vocabulary of the trade. Some of this vocabulary is not readily accessible in everyday language situations.
One aspect to consider about vocational vocabulary is whether there is evidence of shared vocabulary between the trades. Table 8.2 has the words underlays and claddings in Plumbing and in Carpentry. This sharing in the Construction trades is not surprising but the extent of the overlap has yet to be explored in research. Such research would be useful in determining whether there might be a shared vocabulary in trades. Shared vocabulary could be found between various types of Engineering, perhaps, but careful checking is needed to ensure that technical meanings are taken into account in any comparison.
Another point to consider is how members of a possible word family, such as drainlayer, drainlaying, drainlayers occur in a corpus (see the first column in Table 8.2). Debate around the unit of counting for word lists is ongoing (see Nation, 2016 and Chapter 2), but it is important to keep in mind that while individual types can have technical meanings, not all members of a word family will necessarily also be technical in nature. Consider fix and fixings, for example, as an example from Carpentry. Fix is a fairly common word in general English, in the sense of repair or attach something to something, usually a wall. Fixings however occurred only in the trades corpora, and did not occur in Nation’s BNC/COCA lists.
The table also shows a range of compounds in the trades, such as lead-screw and underlay. All three trades include compounds, which presents some interesting problems when it comes to identifying and classifying technical terms. The first problem is whether these words belong with a word family or whether they should go into a list of compounds. Another problem is that some compounds might be hyphenated in one text but not hyphenated in another. Hyphens present their own problems in corpus analysis for word lists (Nation, 2016).