Page 14 - ELG1601 May Issue 436

ELEuropean Union
Page 14 May 2016
The art of analysing a corpus

e’ve been working
as EAP tutors at the
WUniversity of the Arts
London (UAL) for many years,
supporting a range of degree David King and Helen Hickey of the University of the Arts London Language Centre reveal
courses in a variety of disciplines
such as illustration, materials the patterns emerging from their academic corpora
design and fashion. One of the Grammar: Pronouns
challenges is selecting appropri-
ate materials. Although there are Many EAP coursebooks and reference materials advise
many EAP coursebooks, we’ve Lexis: Creating corpus-informed materials against the use of personal pronouns, and this indeed
found that ‘creative’ language (the may be appropriate for many disciplines. Our analysis of
language of and for art and design) As we describe below, We do this according to 3. students require our English for Graphic Design Corpus shows, however,
is conspicuous by its absence. there are a number of four core principles: repeated exposure to that ‘you’ is a keyword. Supplementary analysis of writ-
We both focused on develop- criteria which are useful 1. language should be key lexical items ten materials for a course on graphic branding revealed
ing and analysing corpora for in drawing up a word list presented within texts 4. students need to be both ‘you’ and ‘your’ to be within the top 25 keywords.
our MAs, so we were aware of from a corpus. Once you that are as ‘authentic’ scaffolded towards Initially, this was surprising, but discipline tutors con-
the potential value of corpus- have this you can use as possible actively using the target firmed that a reasonable use of first- and second-person
informed approaches. Fortunately, it to create classroom 2. authenticity in itself is not vocabulary personal pronouns and possessive adjectives was con-
our managers at UAL have allo- materials. enough – texts should sidered an appropriate feature of speaking and writing
cated us the time to compile a For example, in our also be as relevant as In our Presessional within art and design.
number of corpora, including a Presessional Core possible to the students’ Core Materials the vocab- At UAL students’ practices as artists and designers
corpus of learner writing, the UAL Materials we explic- immediate and future ulary items are embedded are often seen as an extension of the self, so personal
Learner Corpus (UALLC) and itly include vocabulary academic needs, follow- within contexts relevant pronouns such as ‘I’, ‘me’ and ‘my’ are not viewed as
a corpus of English for graphic areas and items which ing Brown’s key principle to art and design stu- something to be avoided. In fact, avoiding the use
design (EGDC). our analysis reveals to of ‘situated cognition’ dents, and are frequently of these pronouns and possessives would, in many
Here, we share some of the be important. (Brown et al., 1989) repeated. instances, detract from the expression of the personal.
stand-out results.

Grammar: Nominalisation Lexis: Keyword lists greater value to students. Below, for courses. Here is the list of keywords
example, are the top ten keywords from a subcorpus on gender, taken
Although nominalisation frequently appears in EAP When deciding which vocabulary
from our English for Graphic Design from texts for a taught module on
coursebooks and reference books, there has been items to focus on, lists of high-fre- corpus. Cultural and Historical Studies.
extremely limited analysis of this within ‘creative’ art quency words may seem appealing,
and design language but this is often not the case. The Rank English for Graphic Design Rank Keywords for gender
Analysis of our own students’ writing in the UALLC table below shows the most fre-
1 design (inc designer/s) 1 women
shows an ability to replicate explicitly taught nomi- quently occurring words in two 2 graphic 2 fashion
nalisations (following the pattern of ‘the noun of the corpora: British Academic Written
3 visual 3 gaze
noun’; e.g. …the denotation of the image… or …the English (BAWE) and our English for 4 typography 4 media
connotation of the colour…). However, our analysis Graphic Design (EGDC).
5 art 5 beauty
also highlighted a lack of manipulation and an absence 6 colour 6. salon
of an extended range of nominalised phrasing. As a Rank BAWE EGDC
7 communication 7 magazines
result, we have developed materials which explicitly 1 the the 8 elements 8 dress
introduce students to this feature of academic writing 2 of of
9 printing 9 body
within the discourses of art and design. 3 and and 10 book 10 gender
4 to to
5 in a Even if you haven’t got a full A useful software tool for creat-
6 a in
academic corpus to work from, a ing a keyword list is AntConc, a
7
is
is
Who reads the 8 that that keyword analysis can be a useful freeware suite of programmes that
tool in creating a word list for indi-
comes with YouTube tutorials and
as
9
design
ELgazette? 10 On reflection, it is obvious that
vidual courses or even elements of
pdf how-to instructions.
be
as
functional language dominates lists
a material representation, as
of word frequency. However, where HIGH DEFINITION image (n)
in a photograph or a picture
content words do appear (‘design’
within English for Graphic Design), (UAL Learner Corpus)
these are closely connected to the
domain of the corpus.
In creating a word list, we have
found that keyword analysis is a
more useful tool. This is a compu-
tation of words that are unusually
frequent within a body of texts, so
‘design’ is unusually frequent in
texts about graphic design. Key- Courtesy Ana Escobar
word analysis of relevant texts
reveals language of potentially much
Lexis: Semantics Word Primary Semantic Field
Taking vocabulary items from the AWL British Academic Written English UAL corpora
Academic Word List (Coxhead,
2000) to inform EAP classroom input found (vb) discovered, as in a (vb) physically located, as in
and materials design can be useful, scientific discovery a space or within a work of
but this is not always as straightfor- art (UALLC)
ward as it first seems. The following image (n) a look, as in an illusory appear- (n) a material representa-
very small random sample of words, ance or an attempt to conform tion, as in a photograph or a
taken from the AWL, exemplifies to an expected appearance picture (UALLC)
properties and characteristics that vision (n) sight, as in physiology, (n) seeing, as in a feature of
need to be considered when teach- a sensory experience gaze and power relations
ing vocabulary.

Lexis: Mid-frequency to our students via our Presessional These figures show exposure in
Core Materials (see Lexis: Creating reading texts and do not include
Mid-frequency vocabulary generally corpus-informed materials above). how often students would have
refers to lexical items occurring in The table below provides a sample of heard the words used in listening
the top 4,000 – 8,000 words (k-4 to such mid-frequency words and how exercises, lectures and the class-
k-8). Recent research has highlighted often they occur within our materials. room.
the lack of a principled approach to It is reasonable to assume then
teaching this area of vocabulary. k-4 metaphor (metaphorical) 10 that students’ ‘true’ exposure would
400 officials in 48 countries writing at UAL reveals that slightly k-5 celebrity 17 have far exceeded this.
Our initial analysis of learners’
signifier
18
The lesson we have taken from
receive their copy every month over 14 per cent of students’ writ- signified 17 this is that our principled approach
ten texts contained vocabulary from
– courtesy of the British Council the mid-frequency range. Encourag- k-6 denotation (denote) 5 to selecting which vocabulary
mythology (myth)

12
to teach, and how we teach it, is
ingly, a significant number of these
items had been explicitly introduced k-7 juxtaposition 5 showing indications of success.