Vocabulary and English for Specific Purposes Research - Averil Coxhead 2018
Construction trade: Plumbing and tutor decision tasks
Vocabulary in the trades
The focus of this section is drawing on experts to identify specialised vocabulary in the trade, once a written corpus analysis had been carried out and general English words with no specialised meanings had been removed from the data set (except for items we wanted to check). As part of the selection processes for deciding on lexical items for a Plumbing list (Coxhead & Demecheleer, under review), a tutor decision task was developed. Chapter 2 highlighted some of the main concerns around consulting experts for insights into technical or specialised vocabulary, particularly those raised by Schmitt (2010) who notes that experts tend not to agree. For the LATTE project, we were not looking for agreement between experts per se, although complete agreement would have been a particularly sweet outcome of the process. Rather, we were conscious that the Applied Linguistics researchers were not well versed in the vocabulary of Plumbing, and we were concerned that lexical items might be overlooked in our quest for specialised vocabulary. Guidance from experts, and in particular, experts who were involved in teaching Plumbing, was used therefore to compare with results from our non-expert analysis of specialised vocabulary. This process began with an analysis of the written Plumbing corpus, using Nation’s (2016) BNC/COCA lists. The results of this analysis showed that a large number of words from the first 3,000 or high frequency lists from Nation occurred in the corpus (over 86%). Examples of these words include building, gas, pipe, pressure, drain and discharge. These lexical items were problematic, since they are highly likely to occur in general texts as well. Therefore, we needed to decide which words from these lists would be candidates for including in a word list of Plumbing. At this point, it is important to remember that the list being developed was for Plumbing only, not for other trades. Some of these lexical items might well also be specialised vocabulary for other trades, Carpentry in particular.
To start this process, we selected all of the words in the 3,000 BNC lists which appeared ten times or more in the professional writing corpus. This frequency cut off was decided because it provided a large number of samples to work from and the frequency figures dropped quickly from under ten instances down to one (and there was a long tail). For these high frequency items, two researchers worked independently and rated all the items as general or technical. They used the Plumbing corpus, dictionaries and websites to check the general English items. Any items which appeared to be both general and specific, such as water, were set aside for inclusion in the expert tutor decision task, so that professional judgements could be used to guide word list selection. Items from the 4,000 and other lists which occurred more than ten times in the corpus were considerably less problematic for selection, because they were clearly specialised in nature, but these items were also set aside for the tutor decision task. Some lexical items were completely new to the research team.
A total of 815 items were selected for the tutor decision task. The first part of the decision task was a ’warm-up task’ of 50 items which all the tutors did so that we could compare their responses to the same words. The remaining 765 words were divided into two separate lists to make the decision task more manageable for the tutors. Table 8.5 shows some examples of the warm-up task items in the left column. The middle column asks for ranking, whereby 2 is for items that they see as technical, 1 is for items that are related to Plumbing but not so technical and 0 for items which are not related to Plumbing at all. The column on the right asks tutors to note whether they feel they need to teach this word to students in their Plumbing classes.
The tutors took this task very seriously, and took time to discuss the meanings of words and what decisions they make about teaching lexical items in class. Three tutors did this task. One tutor ranked all the words in both tasks, because he was keen to discuss vocabulary in Plumbing and to see all the words that had been selected. The two other tutors did one task each.
Did they agree on their rankings? As we had expected, the level of disagreement was much high than the level of agreement on technical words (about 13% agreement), but the discussions and selections all helped provide insight into pedagogical considerations and technical vocabulary in Plumbing. All three tutors agreed on the word shower as a technical word in this field, for example. The tutor who taught the entry-level classes had 68% agreement with the Applied Linguistics researchers on the technicality of words. Decisions on teaching appeared to be made on both personal and professional grounds. An example of a professional decision would be whether the tutors thought the learners might already know a word. If they thought the learners would know a word already, then the word was not given a technical rating. Clearly, if the tutors taught different levels of Plumbing or had different learners in mind, the ratings would be different. Similar factors affected decisions the tutors made about whether to teach a word. Interestingly, the tutors discussed multi-word units as well as single words in their rankings for teaching. For example, if a word was central to a task in Plumbing, then they would discuss whether the lexis would be taught as part of teaching that task. An example is measure an angle, where the target word in the decisions task was angle. Insights such as this are particularly useful for selecting items for word lists and for uncovering more about the nature of vocabulary in the trades. For more on this study and the resulting pedagogical word list, see Coxhead and Demecheleer (under review).
Table 8.5 Warm-up items for the tutor task
The pedagogical word lists have been translated into Tongan, on the basis that trades education is English medium and uses English textbooks (Coxhead, Parkinson & Tu’amoheloa, under review). This research drew on a Pasifika research methodology called Talanoa (Vaioleti, 2006), which is a culturally informed approach to gathering data based on developing warm and reciprocal relationships through face-to-face communication between participants and researchers. Many of the high frequency words in the Carpentry list have Tongan equivalents, such as asphalt/valitā and ceiling/’aofi, but some English words have been adopted as loan words (or have been Tonganised), for example, wire/uaea, mortar/mota, beam/pimi. Dwang(s), however, requires paraphrasing into Tongan: papa pātini pe nōkingi ’o ha alangafale/horizontal bracing pieces between house frame.
The next section looks at the Engineering trades, Automotive Engineering and Fabrication, from the LATTE project. These two areas provide examples and discussion points for specialised vocabulary using qualitative and quantitative data.