university 4/n (45)
chapter?
Vocabulary use in classroom
teaching and textbooks
?.?Introduction
The description of vocabulary use in university contexts is an essential prerequi-
site to the development of effective teaching materials and approaches.There are
many important research questions about word use in university language.For
example,how many words would a student need to know to read a typical univer-
sity textbook?How many to understand a typical university lecture?Do different
academic disciplines use the same words?If not,how much overlap is there across
disciplines?Do some disciplines use a greater range of different words?Are some
words common in everyday use and also common in academic language?What
proportion of a typical academic text is made up of those common words?
Several studies in the past have investigated such questions.Nation and Waring
(1997)provide a thorough survey of previous research on vocabulary size and text
coverage.For example,they cite a study by Goulden,Nation,and Read(1990),
who found that a university graduate will understand about 20,000‘word families’.
Learners with a much smaller vocabulary size can be fairly successful reading many
texts.For example,Hirsh and Nation(1992)found that the 2,000 most common
word families would provide 90%coverage of a corpus of teenage novels.
Studies like these are based on word lists that identify the most important
words in different domains.The most famous of these is the General Service List
(GSL;West 1953),which contains 2,000 common words,identified from analysis
of a 5-million-word general written corpus.The GSL also includes detailed infor-
mation not found in most other word lists,such as percentage figures for different
meanings and different part-of-speech uses of each headword.
Two other word lists have been developed specifically for academic applica-
tions.The University Word List(UWL;Xue&Nation 1984)includes 836 word
families generally common in written academic texts but not included in the GSL.
The Academic Word List(AWL;Coxhead 2000)is similar to the UWL,but it is
derived from a more comprehensive corpus analysis considering the frequency
of academic words,their‘range’(distribution across texts from several different
subdisciplines),and their restriction to academic rather than general texts.The??University Language
AWL,with only 570 word families,achieves comparable(or even better)coverage
of most academic texts than the UWL.
Resources like these have been employed to investigate the number of words
required to understand written texts from different registers and different aca-
demic disciplines.Similar approaches have been applied to spoken texts.For ex-
ample,McCarthy and Carter(1997)describe how vocabulary use in spoken texts
is quite different from what a learner will normally encounter in writing,while
Adolphs and Schmitt(2003)show that normal conversations employ a much
wider set of vocabulary than previously expected.
The present chapter adopts a similar approach to the Adolphs and Schmitt
study,comparing the word use patterns among spoken and written university reg-
isters and academic disciplines(rather than identifying lists of the most important
words).Thus,the following descriptions focus on the distribution of all words
in the T2K-SWAL corpus,comparing the patterns of use for words with differ-
ent distributional profiles(with respect to frequency and range)and for words
functioning in different part-of-speech categories.
Specifically,the chapter describes vocabulary use in classroom teaching and
textbooks from three major perspectives:
–The breakdown of words by frequency level for each register:How many dif-
ferent words are used in each register?How many of these are high-frequency
words and how many are low-frequency,specialized words?
–The breakdown of words by part of speech.For example,how many different
words are nouns,verbs,adjectives,etc.?
–The use of vocabulary in different disciplines.For example,do disciplines
differ in their reliance on specialized vocabulary and how many words are
restricted to a single discipline?
Exhaustive word lists are not included in the present book.However,interested
readers are referred to Biber et al.(2004,Appendix B)for a list of all words in the
T2K-SWAL Corpus.(The monograph can also be accessed on-line,at www.ets.
org/ell/research/toeflmonograph.html.)The sub-lists in that report are organized
by frequency level,and they also distinguish among the words that are found pri-
marily in speech versus those found primarily in writing versus words that are
common in both modes.
?.?A note on methodology
One key research issue for vocabulary analyses is to decide what to count as a
word.In the present case,the analyses were based on‘lemmas’:the base form for
each word,disregarding inflectional morphemes.For example,eat,eats,ate,eating,Chapter 3.Vocabulary use in classroom teaching and textbooks??
and eaten are all realizations of a single lemma:EAT.These inflected word forms
all express the same core meaning associated with the verb lemma EAT.These
inflected variants are thus all treated as realizations of a single vocabulary item in
the word counts.(See Appendix A for a fuller discussion of the methods used to
identify lemmas.)
For the vocabulary analyses,the frequency of each lemma was counted in each
register of the T2K-SWAL Corpus.However,registers are not equally well repre-
sented in the corpus.For example,the sub-corpus for classroom teaching contains
1.248 million words,while the sub-corpus for textbooks includes only.76 million
words(see Table 2.1 in Chapter 2).To compensate for these differences,it is nec-
essary to‘normalize’all raw frequency counts to a rate of occurrence per 1 million
words.These normalized rates of occurrence can then be compared directly across
registers.
For example,the lemma work as a noun occurs 1095 times in the spoken texts
of the T2K-SWAL corpus,and the total word count for the spoken part of the
corpus is c.1,665,000 words.Thus,the normed rate of occurrence for work/n in
the spoken mode is:
1,095/1,665,000*1,000,000=657.15 times per million words
There are two other major methodological considerations that should be borne
in mind for any quantitative study of vocabulary:the representativeness of the
corpora(including the actual topics covered in the corpus),and the problems en-
countered in comparing vocabulary distributions across corpora of different sizes
(because word type distributions are not linear relationships).Because these are
both relatively complex methodological issues,I deal with them in some detail in
Appendix B.That appendix includes the results of a series of methodological ex-
periments,testing the effects of corpus size on the apparent vocabulary diversity.
In sum,those experiments show that a small corpus will(misleadingly)seem to
use a larger stock of different words than a large corpus,because words tend to
be repeated in a larger corpus.The appendix introduces a formula to‘normalize’
word type counts to a common basis(per one million words),together with ex-
periments that illustrate how this formula enables comparisons across corpora of
different size.
This normalization procedure is used for the findings presented in Section
3.3 below,which compares the patterns of word use for two registers:classroom
teaching and textbooks.That section also discusses differences in word use across
the academic disciplines within each register.These results should be considered
preliminary,because they are based on a corpus that is small for the purposes of
vocabulary investigations,and because the norming of word type counts provides
only an estimate of the non-linear relationship between word types and corpus
size(see Appendix B).It would be inappropriate under these circumstances to??University Language
30000
25000
20000
15000
10000
5000
0
NmberufowordytespNO(Torn
alizmd)e1–20
21–200
>200
Classroom teaching Textbooks
Figure 3.1 Number of word types at three frequency levels(rates per million words)
do detailed analysis of individual words.However,Section 3.3 shows that there
are large differences in the general patterns of vocabulary use across university
registers and disciplines,and the methods applied here are more than adequate
for describing those major trends.
?.?Vocabulary use in university registers
?.?.?Vocabulary in classroom teaching and textbooks
Classroom teaching and textbooks are similar in their overall purposes and topics.
The primary situational difference between the two is that classroom teaching is
spoken and produced in real time,while textbooks are written and therefore care-
fully planned,revised,and edited.However,it turns out that this situational differ-
ence has a strong influence on word choice:classroom teaching in the T2K-SWAL
Corpus uses c.14,500 different words,while textbooks use c.27,000 different
words.1
Further analysis shows that the greater diversity in word choice in textbooks
is due mostly to the use of specialized vocabulary.Figure 3.1 plots the breakdown
of word types by frequency level:very common words(occurring more than 200
times per million words;e.g.,become,make,great);moderately common words
(occurring between 21 and 200 times per million words;e.g.,afraid,compare,con-
fidence),and rare words(occurring fewer than 20 times per million words;e.g.,
affiliation,buoyancy,commensurate).Chapter 3.Vocabulary use in classroom teaching and textbooks??
cOurcrecnseerpilmionlworsd6000
5000
4000
3000
2000
1000
0
Classroom
teaching
Textbooks
Get Sa
y
Think W
ant See
Thing Oc
cur
Include Co
nt
rol
An
alysis
De
ve
lop
metA
c
tio
n
Figure 3.2 Selected words with especially high frequencies in classroom teaching or text-
books
Both registers are similar in using relatively few high-frequency word types.
But the registers differ dramatically in their reliance on rare word types,with text-
books using a much larger set of these specialized words than classroom teaching.
In contrast,many common words occur with extremely high frequencies in class-
room teaching.Figure 3.2 plots the normed rate of occurrence for some of these
words.For example,verbs such as get,say,think,want,and see all occur well over
2,000 times per million words in classroom teaching,while the noun thing occurs
over 3,000 times per million words in teaching.There are also many words that
occur especially in textbooks,but none of these words occur with extremely high
frequencies.For example,Figure 3.2 shows that some relatively common verbs
(e.g.occur,include)and several relatively common nouns(e.g.,control,analysis,
development,action)occur more often in textbooks,but even the most frequent of
these words occur less than 1,000 times per million words.
Thus,the general picture emerging from Figures 3.1–3.2 is the following:
–Textbooks use a much greater range of different word types than classroom
teaching.
–Textbooks and classroom teaching both use a relatively small set of common
words.
–Many of these common words occur with extremely high frequencies in class-
room teaching.
–In contrast,textbooks use a wide range of word types that occur with low
frequencies.??University Language
Nmberu
fowordytespNO(Tornalizmd)e18000
16000
14000
12000
10000
8000
6000
4000
2000
0
Nouns Verbs Adjs Advs Nouns Verbs Adjs Advs
1–20
21–200
>200
Classroom teaching Textbooks
Figure 3.3 Number of word types at three frequency levels,by grammatical word class
(rates per million words)
Figure 3.3 shows that the majority of the different word types in the T2K-SWAL
Corpus are nouns.This is especially the case for rare word types,although mod-
erately common word types(with frequencies between 21–200 per million words)
show the same pattern.These nouns include some everyday words that generally
are not the normal topic of discussion in teaching or textbooks,such as chalkboard,
cigarette,doorway.However,most of these nouns have more specialized meanings,
like disillusionment,enhancement,globalization,hominid,locus.Textbooks use an
especially large number of different word types functioning as nouns.There are
also a large range of different adjectives used in textbooks(and to a lesser extent
classroom teaching);these are words such as occupational,pediatric,representa-
tional,and sensory.There are a smaller number of different verbs in either register,
and adverbs show the least diversity of the four content word classes.
The following two text excerpts,from a classroom lecture and a textbook
in engineering,illustrate these basic patterns.Both excerpts describe engineering
problems,one concerning flow rate and the other road surfaces.The two excerpts
are also similar in showing a mix of high-frequency and less common words.The
major difference between the two excerpts is in the extent to which they rely on
the different kinds of words.
In the classroom teaching text(Excerpt 3.1)there are several less common
words conveying specialized information(e.g.reaction,molar,reactant,concentra-
tion,differential).However,there is an even greater reliance on common content
words,and many of these are extremely high frequency words,such as:again,look,
see,make,here,gonna,guess,now,mean,have,need,use.It is interesting to noteChapter 3.Vocabulary use in classroom teaching and textbooks??
how these extremely common words are interspersed with technical vocabulary in
the spoken discourse.This seems to reflect the instructor’s awareness of the class-
room audience,using everyday terms to explain what he is doing as he develops a
complex mathematical equation on the blackboard.
In contrast,the excerpt from the engineering textbook(3.2)is noteworthy for
its use of rare words;for example:profilometers,wavenumber,wavelength,ampli-
tude,deviations,deteriorating,bituminous,cycle.
Text Excerpt 3.1:Engineering Classroom Teaching(engceleudli100.txt),
chemical engineering
Relatively frequent content words(occurring more than 400 times per mil-
lion words)are in bold;Less frequent content words(occurring fewer than
200 times per million words)are in italics
So,again taking a look at a specific example,of this,we look at a first order
reaction,the mole balance,for,plug flow reactor is remember,D.F.A.B.V.is equal
to,uh,the rate of reaction.Instead of writing it in terms of molar flow rate,I’m
gonna put everything in terms of concentration.And so,since the molar flow
rate,is the same as,uh concentration times,the volume(that your)flow rate,
I write it this way.And,you see,I’ve sort of made another assumption here
and that is that the volume and the flow rate is not a function of position in a
plug flow reactor.So this equation,in addition,another restriction to this one
is gonna be,note,the volume changes...and the,the negative sign is here
because,A is disappearing.It’s a reactant.And,plugging now into the energy
balance the rate[2 sylls].I mean they don’t have these two boxed equations to
solve simultaneously.And they’re,ordinary differential equations so we’ll need to
use,need to use the homework.
Text Excerpt 3.2:Engineering Textbook(TBMCE3.GVD),mechanical engi-
neering
Relatively frequent content words(occurring more than 400 times per mil-
lion words)are in bold;Less frequent content words(occurring fewer than
200 times per million words)are in italics
Road elevation profiles can be measured either by performing close interval rod
and level surveys or by high speed profilometers.When the PSDs are determined,
plots such as those shown in Figure 5.2 are typically obtained.Although the PSD
of every road section is unique,all roads show the characteristic drop in ampli-
tude with wavenumber.This simply reflects the fact that deviations in the road
surface on the order of hundreds of feet in length may have amplitudes of inches,
whereas those only a few feet in length are normally only fractions of an inch in
amplitude.The general amplitude level of the plot is indicative of the roughness
level with higher amplitudes implying rougher roads.The wavenumber range in??University Language
the figure corresponds to wavelengths of 200 feet(61 m)on the left at 0.005 cy-
cle/foot(0.016 cycle/meter),to about 2 feet(0.6 m)on the right at 0.5 cycle/foot
(1.6 cycles/meter).
Another perspective on vocabulary use is to consider the specialized word types
that are restricted to either speech or writing.Figure 3.4 shows that there are very
few word types used in classroom teaching that are not also used in textbooks.
That is,most of the word types that a student encounters in classroom teaching
are also used in textbooks.In contrast,textbooks rely on many specialized word
types that are not found in the corresponding classroom teaching sessions.2 Here
again we see the greatest degree of specialization for nouns:almost half of the dif-
ferent nouns in the T2K-SWAL Corpus are found only in textbooks.These include
some relatively common words like self,agent,and combination.However,many of
these nouns are rare words with highly specialized meanings,like agglomeration,
chromatography,dialectic,electrode,felony.
Surprisingly,Figure 3.4 shows there are some nouns found exclusively in class-
room teaching.Many of these nouns are everyday words with meanings that are
not normally discussed in a written textbook,e.g.:bagel,bakery,banana,nail,
tourist,parking.There are also colloquial nouns(and other words)used in class-
room teaching that would usually not be considered appropriate in a textbook;for
example:bug,buzz,cop,chump,dude,fluff.3 However,the existence of this small
set of words found exclusively in classroom teaching does not obscure the overall
NmberufowordytespNO(Tornalizmd)eClassroom Teaching ONLY Textbooks ONLY Both registers
Adverbs
Adjectives
Verbs
Nouns
16000
14000
12000
10000
8000
6000
4000
2000
0
Figure 3.4 Distribution of specialized word types–restricted to either classroom teaching
or textbooks(by grammatical class)Chapter 3.Vocabulary use in classroom teaching and textbooks??
pattern:that there is a much greater range of word types used in textbooks than in
classroom teaching,and most of these different words are nouns.
These patterns are mostly a reflection of the different production circum-
stances of classroom teaching versus textbooks.Textbook authors have extensive
time for producing their texts,including the initial writing as well as revision and
editing.In contrast,instructors in classroom teaching usually pre-plan their dis-
course,but the actual spoken text is created as it is produced in the classroom.As a
result,instructors in classroom teaching rely on a relatively small set of words,but
they use a few words with extremely high frequencies.(Their speech also tends
to rely on clauses rather than elaborated noun phrases;see Chapter 4.)In con-
trast,the extended production opportunities for textbook authors allow the use of
a much larger stock of words,including the selection of words with specific rather
than general meanings.Textbook authors are also motivated by a stylistic prefer-
ence for varied vocabulary,rather than using the same word repeatedly(at least
in some academic disciplines–see Section 3.3.2 below).Finally,textbook authors
employ elaborated noun phrase constructions,relying to a large extent on phrases
rather than clauses to convey information(see the discussion of grammatical fea-
tures in Chapter 4).As a result,textbooks show much greater vocabulary diversity
than classroom teaching,with most of that diversity being realized as an extremely
large set of different nouns.
?.?.?Vocabulary across academic disciplines
In addition to the overall differences between classroom teaching and textbooks,
there are also important differences in word use across academic disciplines.Fig-
ure 3.5 plots the number of word types(normalized per 1-million words)in each
discipline.Classroom teaching and textbooks both show the same overall patterns:
Business and engineering have much less diversity in word choice than natural sci-
ence,social science,and humanities.These differences are much more pronounced
in textbooks than in classroom teaching.Humanities textbooks are especially note-
worthy,with an extremely large set of word types.However,social science and
natural science textbooks also use a very large set of different word types.Figure
3.6 shows the breakdown of specialized word types across academic disciplines,
plotting the number of word types found in only a single discipline(i.e.,‘special-
ized’words)versus word types that are used in several disciplines.Similar to Figure
3.5,Figure 3.6 shows a major difference between Business and Engineering,on the
one hand,and the more general Sciences and Humanities(Natural Science,Social
Science,and Humanities)on the other.
In part,these differences seem to reflect the range of subject areas included un-
der each of these academic disciplines.Business and engineering are professional
disciplines,training students in specific skills and methods.As a result,the set of??University Language
Wordytesperp
ilmionlworsd35000
30000
25000
20000
15000
10000
5000
0
Business Engineering Natural Science Social Science Humanities
Classroom teaching Textbooks
Figure 3.5 Number of word types across academic disciplines
Nmberufowo
rdytespNO(Tornalizmd)e20000
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
Bu
siness
ONL
Y
Eng
ineer
ing
ONL
Y
Na
tur
alS
cienc
e
ONL
Y
Social
Scienc
e
ONL
Y
Hu
manitiesO
NL
YM
ultiple
disciplines
Textbooks
Classroom
teaching
Figure 3.6 Distribution of specialized word types–restricted to a single academic disci-
pline(vs.general word types found in multiple disciplines)Chapter 3.Vocabulary use in classroom teaching and textbooks??
topics covered in these general disciplines is somewhat more constrained than in
the sciences and humanities.Natural science might be characterized as a discipline
of discovery,identifying and describing entities that had not been previously con-
sidered.As a result,natural science employs a large set of highly technical words,
like dextrinoid,electrophoresis,and phallotoxins.Most of these words do not have
commonplace synonyms,because they refer to entities,characteristics,or concepts
that are not normally discussed in everyday conversation.In contrast,humanities
and social science textbooks are more likely to deal with aspects of everyday life,
discussing people,events,and social behavior from new perspectives.Humanities
and social sciences employ a very large set of specialized words,but many of these
terms provide a single word for an entity or concept that can easily be discussed in
everyday conversation with a fuller phrase.For example:
ingrate=someone who doesn’t appreciate something
misconception=a really strange idea
pedagogy=a style of teaching
sanctimonious=he thinks he’s“holier-than-thou”
To illustrate the kinds of words preferred in each academic discipline,I carried out
a case study considering all specialized vocabulary beginning with the letter‘A’in
the T2K-SWAL Corpus.Table 3.1 shows the breakdown of these specialized words
across disciplines.
Business and Engineering are similar in having only a few technical terms
that are restricted to that single discipline.These are words like accrual,annuity,
and audit in Business,and absorber,aerodynamics,and attenuate in Engineering.
More commonly,these two disciplines use everyday terms with a specific technical
meaning,resulting in an extremely frequent use of the term;these are words like
account,act,and adjust in Business,and address,arm,and assembly in Engineering.
The pattern of word use in Natural Science is quite different.Table 3.1 lists
a large number of specialized terms found exclusively in this discipline;most of
these words are highly technical in meaning and have no counterpart in everyday
usage.These are words like abscission,acastia,acetylation,achene,acyanogenic,etc.
Humanities and social sciences also have a large number of specialized words
that are restricted to these disciplines.However,as noted above,these words tend
to refer to concepts or entities that could easily be described in normal conversa-
tion with a fuller expression.For example:
altruism=caring about other people
ambivalence=having mixed feelings about something
amoral=doesn’t care about right or wrong
Detailed consideration of these specialized words shows that there are strikingly
different patterns of use across disciplines.Figures 3.5 and 3.6 show a general??University Language
Table 3.1 Specialized vocabulary beginning with the letter‘A’,broken down by academic
discipline
Business
Words found only(or primarily)in Business:
accredited,accrual,adjusted/ing,adversarial,affective/ity,amend,annualized,annuity,
apprentice,arbitration,audit,auditor,averse/ion
Words that are much more common in Business than in other disciplines:
accommodation,account,accounting,accumulated,accuracy,achievement,act(noun),action,
activate,activity,ad,adjust,administer,adopt,advantage,adverse,advertise,advertiser,affect,
affiliation,affirmative,agency,agent,agreement,allege,allocation,allowance,alternative,amend,
amendment,analysis,annual,approach,ask,aspect,assertion,assess/ment,asset,assurance,
attention,attractive/ness
Engineering
Words found only(or primarily)in Engineering:
absorber,adiabatic,aerodynamics,aerospace,algorithm,alloy,analyzer,annular,artifactual,
attenuate/ion,automated,axle
Words found only in Engineering and Science:
absorb,acetone,advection,airflow,algebraic,ammonia,ammonium,analog,angular,anisotropic,
annulus,anthracite,approximately,aqueous,aquifer,asbestos,aspherical,axis
Words that are much more common in Engineering than in other disciplines:
absolute,acceleration,access,accordingly,actual,address,alpha,ambiguous,amplify,amplitude,
analogous,applicability,application,apply,approximate/ion,arbitrary,architecture,arithmetic,
arm,array,arrow,assembly,assignment,associate
Natural Science
Words found only(or primarily)in Natural Science:
abscission,acastia,accretion,acetate,acetylation,achene,actin,acyanogenic,adnation,adsorb,
adsorption,advection,aeration,agaric,albumin,alder,algae,aliele,aliphatic,alkaline,
alleghaniensis,allotropic,allozyme,alluvial,alphape,aluminosilicates,amanita,amatoxin,
amphibole,anaerobic,andesite,anemophilous,anhydrite,anionic,anisotropic,anode,anoxic,
antiparallel,antiquark,antiviral,aperiodic,apetalous,aphyllophorale,apophysis,appendage,
arbuscules,archaebacteria,archean,ardente,arkose,armillaria,ascomycete,asepalous,asteroid,
asthenosphere,autochory,autoclastic,autodeliquescence,axil,axillary
Humanities and Social Science
Words found only(or primarily)in Humanities and/or Social Science:
altruism,ambivalence,amoral,anachronistic,anarchy,ancillary,anecdotal,angel,anger,anguish,
animosity,annexation,antagonism,anthology,antiquity,antithesis,antiwar,anxiety,apathy,
apostle,apostrophe,apprehend,apprehension,archaic,archetype,aristocracy,arousal,arrest,
arrogant,artifice,artisan,artistic,ascetic,ascribe,asocial,aspiration,assailant,assassin,assault,
assimilate,astonishment,astrology,astronomical,asylum,atheism,atonement,atrocity,attendant,
attentive,attribution,atypical,auspicious,austere,authenticity,authoritative,autobiography,
autocratic,autonomy,autopsy,avenge,avert,avid,awfulChapter 3.Vocabulary use in classroom teaching and textbooks??
similarity between Natural Science,Social Science,and Humanities:all three dis-
ciplines have a large number of different word types,including many words re-
stricted to a single discipline.However,this quantitative similarity corresponds to
very different patterns of word use in the disciplines.Natural Science uses a large
stock of technical terms for highly specialized reference:words that refer to entities,
characteristics,and concepts that are not readily discussed in everyday conversa-
tion.In contrast,Humanities and Social Sciences often offer new perspectives on
concepts and entities that are taken from our everyday experience.As a result,the
extensive stock of specialized words in these disciplines are often technical terms
to refer to these everyday experiences.
?.?Conclusion
This chapter has taken a different approach to vocabulary study than most pre-
vious studies of academic discourse,focusing on the diversity of vocabulary and
the frequency distributions of common and rare words across registers(rather
than the use of specific words or the identification of lists of the most common
words).This perspective exposed interesting contrasts between spoken and writ-
ten registers and across disciplines.Overall,classroom teaching was found to use
a relatively small set of different word types,but to rely heavily on a few of those
words,which therefore occur with extremely high frequencies.Textbooks,in con-
trast,were found to use a larger set of different word types(especially different
nouns),but none of those individual word types occur with extremely high fre-
quencies.Such distributions are consistent with the differences between the two
registers with respect to their planning and revising time,as well as the presence
of a face-to-face audience.
The comparison of vocabulary patterns across disciplines highlighted the re-
lationship between word use and subject matter.Business and engineering were
found to have less diversity in word choice,using words that are adapted from
everyday use but have specific technical meanings.Natural science has more di-
versity in vocabulary,reflecting the diversity of its sub-disciplines,and it has more
rarely occurring technical terms that have no everyday counterparts.The human-
ities and social sciences also evidence diverse word choice,consistent with their
diverse subject matter;many words in these disciplines could be summarized with
longer phrases in everyday language,as the disciplines address matters that are
more often part of our everyday experience.
Other perspectives on vocabulary use in academic language are of course
needed.For example,it would be useful to investigate the use of multi-word terms,
such as“hard income measure”in business or“truth value”in philosophy.Never-
theless,even the limited analyses possible with the relatively small sub-corpora in
Vocabulary use in classroom
teaching and textbooks
?.?Introduction
The description of vocabulary use in university contexts is an essential prerequi-
site to the development of effective teaching materials and approaches.There are
many important research questions about word use in university language.For
example,how many words would a student need to know to read a typical univer-
sity textbook?How many to understand a typical university lecture?Do different
academic disciplines use the same words?If not,how much overlap is there across
disciplines?Do some disciplines use a greater range of different words?Are some
words common in everyday use and also common in academic language?What
proportion of a typical academic text is made up of those common words?
Several studies in the past have investigated such questions.Nation and Waring
(1997)provide a thorough survey of previous research on vocabulary size and text
coverage.For example,they cite a study by Goulden,Nation,and Read(1990),
who found that a university graduate will understand about 20,000‘word families’.
Learners with a much smaller vocabulary size can be fairly successful reading many
texts.For example,Hirsh and Nation(1992)found that the 2,000 most common
word families would provide 90%coverage of a corpus of teenage novels.
Studies like these are based on word lists that identify the most important
words in different domains.The most famous of these is the General Service List
(GSL;West 1953),which contains 2,000 common words,identified from analysis
of a 5-million-word general written corpus.The GSL also includes detailed infor-
mation not found in most other word lists,such as percentage figures for different
meanings and different part-of-speech uses of each headword.
Two other word lists have been developed specifically for academic applica-
tions.The University Word List(UWL;Xue&Nation 1984)includes 836 word
families generally common in written academic texts but not included in the GSL.
The Academic Word List(AWL;Coxhead 2000)is similar to the UWL,but it is
derived from a more comprehensive corpus analysis considering the frequency
of academic words,their‘range’(distribution across texts from several different
subdisciplines),and their restriction to academic rather than general texts.The??University Language
AWL,with only 570 word families,achieves comparable(or even better)coverage
of most academic texts than the UWL.
Resources like these have been employed to investigate the number of words
required to understand written texts from different registers and different aca-
demic disciplines.Similar approaches have been applied to spoken texts.For ex-
ample,McCarthy and Carter(1997)describe how vocabulary use in spoken texts
is quite different from what a learner will normally encounter in writing,while
Adolphs and Schmitt(2003)show that normal conversations employ a much
wider set of vocabulary than previously expected.
The present chapter adopts a similar approach to the Adolphs and Schmitt
study,comparing the word use patterns among spoken and written university reg-
isters and academic disciplines(rather than identifying lists of the most important
words).Thus,the following descriptions focus on the distribution of all words
in the T2K-SWAL corpus,comparing the patterns of use for words with differ-
ent distributional profiles(with respect to frequency and range)and for words
functioning in different part-of-speech categories.
Specifically,the chapter describes vocabulary use in classroom teaching and
textbooks from three major perspectives:
–The breakdown of words by frequency level for each register:How many dif-
ferent words are used in each register?How many of these are high-frequency
words and how many are low-frequency,specialized words?
–The breakdown of words by part of speech.For example,how many different
words are nouns,verbs,adjectives,etc.?
–The use of vocabulary in different disciplines.For example,do disciplines
differ in their reliance on specialized vocabulary and how many words are
restricted to a single discipline?
Exhaustive word lists are not included in the present book.However,interested
readers are referred to Biber et al.(2004,Appendix B)for a list of all words in the
T2K-SWAL Corpus.(The monograph can also be accessed on-line,at www.ets.
org/ell/research/toeflmonograph.html.)The sub-lists in that report are organized
by frequency level,and they also distinguish among the words that are found pri-
marily in speech versus those found primarily in writing versus words that are
common in both modes.
?.?A note on methodology
One key research issue for vocabulary analyses is to decide what to count as a
word.In the present case,the analyses were based on‘lemmas’:the base form for
each word,disregarding inflectional morphemes.For example,eat,eats,ate,eating,Chapter 3.Vocabulary use in classroom teaching and textbooks??
and eaten are all realizations of a single lemma:EAT.These inflected word forms
all express the same core meaning associated with the verb lemma EAT.These
inflected variants are thus all treated as realizations of a single vocabulary item in
the word counts.(See Appendix A for a fuller discussion of the methods used to
identify lemmas.)
For the vocabulary analyses,the frequency of each lemma was counted in each
register of the T2K-SWAL Corpus.However,registers are not equally well repre-
sented in the corpus.For example,the sub-corpus for classroom teaching contains
1.248 million words,while the sub-corpus for textbooks includes only.76 million
words(see Table 2.1 in Chapter 2).To compensate for these differences,it is nec-
essary to‘normalize’all raw frequency counts to a rate of occurrence per 1 million
words.These normalized rates of occurrence can then be compared directly across
registers.
For example,the lemma work as a noun occurs 1095 times in the spoken texts
of the T2K-SWAL corpus,and the total word count for the spoken part of the
corpus is c.1,665,000 words.Thus,the normed rate of occurrence for work/n in
the spoken mode is:
1,095/1,665,000*1,000,000=657.15 times per million words
There are two other major methodological considerations that should be borne
in mind for any quantitative study of vocabulary:the representativeness of the
corpora(including the actual topics covered in the corpus),and the problems en-
countered in comparing vocabulary distributions across corpora of different sizes
(because word type distributions are not linear relationships).Because these are
both relatively complex methodological issues,I deal with them in some detail in
Appendix B.That appendix includes the results of a series of methodological ex-
periments,testing the effects of corpus size on the apparent vocabulary diversity.
In sum,those experiments show that a small corpus will(misleadingly)seem to
use a larger stock of different words than a large corpus,because words tend to
be repeated in a larger corpus.The appendix introduces a formula to‘normalize’
word type counts to a common basis(per one million words),together with ex-
periments that illustrate how this formula enables comparisons across corpora of
different size.
This normalization procedure is used for the findings presented in Section
3.3 below,which compares the patterns of word use for two registers:classroom
teaching and textbooks.That section also discusses differences in word use across
the academic disciplines within each register.These results should be considered
preliminary,because they are based on a corpus that is small for the purposes of
vocabulary investigations,and because the norming of word type counts provides
only an estimate of the non-linear relationship between word types and corpus
size(see Appendix B).It would be inappropriate under these circumstances to??University Language
30000
25000
20000
15000
10000
5000
0
NmberufowordytespNO(Torn
alizmd)e1–20
21–200
>200
Classroom teaching Textbooks
Figure 3.1 Number of word types at three frequency levels(rates per million words)
do detailed analysis of individual words.However,Section 3.3 shows that there
are large differences in the general patterns of vocabulary use across university
registers and disciplines,and the methods applied here are more than adequate
for describing those major trends.
?.?Vocabulary use in university registers
?.?.?Vocabulary in classroom teaching and textbooks
Classroom teaching and textbooks are similar in their overall purposes and topics.
The primary situational difference between the two is that classroom teaching is
spoken and produced in real time,while textbooks are written and therefore care-
fully planned,revised,and edited.However,it turns out that this situational differ-
ence has a strong influence on word choice:classroom teaching in the T2K-SWAL
Corpus uses c.14,500 different words,while textbooks use c.27,000 different
words.1
Further analysis shows that the greater diversity in word choice in textbooks
is due mostly to the use of specialized vocabulary.Figure 3.1 plots the breakdown
of word types by frequency level:very common words(occurring more than 200
times per million words;e.g.,become,make,great);moderately common words
(occurring between 21 and 200 times per million words;e.g.,afraid,compare,con-
fidence),and rare words(occurring fewer than 20 times per million words;e.g.,
affiliation,buoyancy,commensurate).Chapter 3.Vocabulary use in classroom teaching and textbooks??
cOurcrecnseerpilmionlworsd6000
5000
4000
3000
2000
1000
0
Classroom
teaching
Textbooks
Get Sa
y
Think W
ant See
Thing Oc
cur
Include Co
nt
rol
An
alysis
De
ve
lop
metA
c
tio
n
Figure 3.2 Selected words with especially high frequencies in classroom teaching or text-
books
Both registers are similar in using relatively few high-frequency word types.
But the registers differ dramatically in their reliance on rare word types,with text-
books using a much larger set of these specialized words than classroom teaching.
In contrast,many common words occur with extremely high frequencies in class-
room teaching.Figure 3.2 plots the normed rate of occurrence for some of these
words.For example,verbs such as get,say,think,want,and see all occur well over
2,000 times per million words in classroom teaching,while the noun thing occurs
over 3,000 times per million words in teaching.There are also many words that
occur especially in textbooks,but none of these words occur with extremely high
frequencies.For example,Figure 3.2 shows that some relatively common verbs
(e.g.occur,include)and several relatively common nouns(e.g.,control,analysis,
development,action)occur more often in textbooks,but even the most frequent of
these words occur less than 1,000 times per million words.
Thus,the general picture emerging from Figures 3.1–3.2 is the following:
–Textbooks use a much greater range of different word types than classroom
teaching.
–Textbooks and classroom teaching both use a relatively small set of common
words.
–Many of these common words occur with extremely high frequencies in class-
room teaching.
–In contrast,textbooks use a wide range of word types that occur with low
frequencies.??University Language
Nmberu
fowordytespNO(Tornalizmd)e18000
16000
14000
12000
10000
8000
6000
4000
2000
0
Nouns Verbs Adjs Advs Nouns Verbs Adjs Advs
1–20
21–200
>200
Classroom teaching Textbooks
Figure 3.3 Number of word types at three frequency levels,by grammatical word class
(rates per million words)
Figure 3.3 shows that the majority of the different word types in the T2K-SWAL
Corpus are nouns.This is especially the case for rare word types,although mod-
erately common word types(with frequencies between 21–200 per million words)
show the same pattern.These nouns include some everyday words that generally
are not the normal topic of discussion in teaching or textbooks,such as chalkboard,
cigarette,doorway.However,most of these nouns have more specialized meanings,
like disillusionment,enhancement,globalization,hominid,locus.Textbooks use an
especially large number of different word types functioning as nouns.There are
also a large range of different adjectives used in textbooks(and to a lesser extent
classroom teaching);these are words such as occupational,pediatric,representa-
tional,and sensory.There are a smaller number of different verbs in either register,
and adverbs show the least diversity of the four content word classes.
The following two text excerpts,from a classroom lecture and a textbook
in engineering,illustrate these basic patterns.Both excerpts describe engineering
problems,one concerning flow rate and the other road surfaces.The two excerpts
are also similar in showing a mix of high-frequency and less common words.The
major difference between the two excerpts is in the extent to which they rely on
the different kinds of words.
In the classroom teaching text(Excerpt 3.1)there are several less common
words conveying specialized information(e.g.reaction,molar,reactant,concentra-
tion,differential).However,there is an even greater reliance on common content
words,and many of these are extremely high frequency words,such as:again,look,
see,make,here,gonna,guess,now,mean,have,need,use.It is interesting to noteChapter 3.Vocabulary use in classroom teaching and textbooks??
how these extremely common words are interspersed with technical vocabulary in
the spoken discourse.This seems to reflect the instructor’s awareness of the class-
room audience,using everyday terms to explain what he is doing as he develops a
complex mathematical equation on the blackboard.
In contrast,the excerpt from the engineering textbook(3.2)is noteworthy for
its use of rare words;for example:profilometers,wavenumber,wavelength,ampli-
tude,deviations,deteriorating,bituminous,cycle.
Text Excerpt 3.1:Engineering Classroom Teaching(engceleudli100.txt),
chemical engineering
Relatively frequent content words(occurring more than 400 times per mil-
lion words)are in bold;Less frequent content words(occurring fewer than
200 times per million words)are in italics
So,again taking a look at a specific example,of this,we look at a first order
reaction,the mole balance,for,plug flow reactor is remember,D.F.A.B.V.is equal
to,uh,the rate of reaction.Instead of writing it in terms of molar flow rate,I’m
gonna put everything in terms of concentration.And so,since the molar flow
rate,is the same as,uh concentration times,the volume(that your)flow rate,
I write it this way.And,you see,I’ve sort of made another assumption here
and that is that the volume and the flow rate is not a function of position in a
plug flow reactor.So this equation,in addition,another restriction to this one
is gonna be,note,the volume changes...and the,the negative sign is here
because,A is disappearing.It’s a reactant.And,plugging now into the energy
balance the rate[2 sylls].I mean they don’t have these two boxed equations to
solve simultaneously.And they’re,ordinary differential equations so we’ll need to
use,need to use the homework.
Text Excerpt 3.2:Engineering Textbook(TBMCE3.GVD),mechanical engi-
neering
Relatively frequent content words(occurring more than 400 times per mil-
lion words)are in bold;Less frequent content words(occurring fewer than
200 times per million words)are in italics
Road elevation profiles can be measured either by performing close interval rod
and level surveys or by high speed profilometers.When the PSDs are determined,
plots such as those shown in Figure 5.2 are typically obtained.Although the PSD
of every road section is unique,all roads show the characteristic drop in ampli-
tude with wavenumber.This simply reflects the fact that deviations in the road
surface on the order of hundreds of feet in length may have amplitudes of inches,
whereas those only a few feet in length are normally only fractions of an inch in
amplitude.The general amplitude level of the plot is indicative of the roughness
level with higher amplitudes implying rougher roads.The wavenumber range in??University Language
the figure corresponds to wavelengths of 200 feet(61 m)on the left at 0.005 cy-
cle/foot(0.016 cycle/meter),to about 2 feet(0.6 m)on the right at 0.5 cycle/foot
(1.6 cycles/meter).
Another perspective on vocabulary use is to consider the specialized word types
that are restricted to either speech or writing.Figure 3.4 shows that there are very
few word types used in classroom teaching that are not also used in textbooks.
That is,most of the word types that a student encounters in classroom teaching
are also used in textbooks.In contrast,textbooks rely on many specialized word
types that are not found in the corresponding classroom teaching sessions.2 Here
again we see the greatest degree of specialization for nouns:almost half of the dif-
ferent nouns in the T2K-SWAL Corpus are found only in textbooks.These include
some relatively common words like self,agent,and combination.However,many of
these nouns are rare words with highly specialized meanings,like agglomeration,
chromatography,dialectic,electrode,felony.
Surprisingly,Figure 3.4 shows there are some nouns found exclusively in class-
room teaching.Many of these nouns are everyday words with meanings that are
not normally discussed in a written textbook,e.g.:bagel,bakery,banana,nail,
tourist,parking.There are also colloquial nouns(and other words)used in class-
room teaching that would usually not be considered appropriate in a textbook;for
example:bug,buzz,cop,chump,dude,fluff.3 However,the existence of this small
set of words found exclusively in classroom teaching does not obscure the overall
NmberufowordytespNO(Tornalizmd)eClassroom Teaching ONLY Textbooks ONLY Both registers
Adverbs
Adjectives
Verbs
Nouns
16000
14000
12000
10000
8000
6000
4000
2000
0
Figure 3.4 Distribution of specialized word types–restricted to either classroom teaching
or textbooks(by grammatical class)Chapter 3.Vocabulary use in classroom teaching and textbooks??
pattern:that there is a much greater range of word types used in textbooks than in
classroom teaching,and most of these different words are nouns.
These patterns are mostly a reflection of the different production circum-
stances of classroom teaching versus textbooks.Textbook authors have extensive
time for producing their texts,including the initial writing as well as revision and
editing.In contrast,instructors in classroom teaching usually pre-plan their dis-
course,but the actual spoken text is created as it is produced in the classroom.As a
result,instructors in classroom teaching rely on a relatively small set of words,but
they use a few words with extremely high frequencies.(Their speech also tends
to rely on clauses rather than elaborated noun phrases;see Chapter 4.)In con-
trast,the extended production opportunities for textbook authors allow the use of
a much larger stock of words,including the selection of words with specific rather
than general meanings.Textbook authors are also motivated by a stylistic prefer-
ence for varied vocabulary,rather than using the same word repeatedly(at least
in some academic disciplines–see Section 3.3.2 below).Finally,textbook authors
employ elaborated noun phrase constructions,relying to a large extent on phrases
rather than clauses to convey information(see the discussion of grammatical fea-
tures in Chapter 4).As a result,textbooks show much greater vocabulary diversity
than classroom teaching,with most of that diversity being realized as an extremely
large set of different nouns.
?.?.?Vocabulary across academic disciplines
In addition to the overall differences between classroom teaching and textbooks,
there are also important differences in word use across academic disciplines.Fig-
ure 3.5 plots the number of word types(normalized per 1-million words)in each
discipline.Classroom teaching and textbooks both show the same overall patterns:
Business and engineering have much less diversity in word choice than natural sci-
ence,social science,and humanities.These differences are much more pronounced
in textbooks than in classroom teaching.Humanities textbooks are especially note-
worthy,with an extremely large set of word types.However,social science and
natural science textbooks also use a very large set of different word types.Figure
3.6 shows the breakdown of specialized word types across academic disciplines,
plotting the number of word types found in only a single discipline(i.e.,‘special-
ized’words)versus word types that are used in several disciplines.Similar to Figure
3.5,Figure 3.6 shows a major difference between Business and Engineering,on the
one hand,and the more general Sciences and Humanities(Natural Science,Social
Science,and Humanities)on the other.
In part,these differences seem to reflect the range of subject areas included un-
der each of these academic disciplines.Business and engineering are professional
disciplines,training students in specific skills and methods.As a result,the set of??University Language
Wordytesperp
ilmionlworsd35000
30000
25000
20000
15000
10000
5000
0
Business Engineering Natural Science Social Science Humanities
Classroom teaching Textbooks
Figure 3.5 Number of word types across academic disciplines
Nmberufowo
rdytespNO(Tornalizmd)e20000
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
Bu
siness
ONL
Y
Eng
ineer
ing
ONL
Y
Na
tur
alS
cienc
e
ONL
Y
Social
Scienc
e
ONL
Y
Hu
manitiesO
NL
YM
ultiple
disciplines
Textbooks
Classroom
teaching
Figure 3.6 Distribution of specialized word types–restricted to a single academic disci-
pline(vs.general word types found in multiple disciplines)Chapter 3.Vocabulary use in classroom teaching and textbooks??
topics covered in these general disciplines is somewhat more constrained than in
the sciences and humanities.Natural science might be characterized as a discipline
of discovery,identifying and describing entities that had not been previously con-
sidered.As a result,natural science employs a large set of highly technical words,
like dextrinoid,electrophoresis,and phallotoxins.Most of these words do not have
commonplace synonyms,because they refer to entities,characteristics,or concepts
that are not normally discussed in everyday conversation.In contrast,humanities
and social science textbooks are more likely to deal with aspects of everyday life,
discussing people,events,and social behavior from new perspectives.Humanities
and social sciences employ a very large set of specialized words,but many of these
terms provide a single word for an entity or concept that can easily be discussed in
everyday conversation with a fuller phrase.For example:
ingrate=someone who doesn’t appreciate something
misconception=a really strange idea
pedagogy=a style of teaching
sanctimonious=he thinks he’s“holier-than-thou”
To illustrate the kinds of words preferred in each academic discipline,I carried out
a case study considering all specialized vocabulary beginning with the letter‘A’in
the T2K-SWAL Corpus.Table 3.1 shows the breakdown of these specialized words
across disciplines.
Business and Engineering are similar in having only a few technical terms
that are restricted to that single discipline.These are words like accrual,annuity,
and audit in Business,and absorber,aerodynamics,and attenuate in Engineering.
More commonly,these two disciplines use everyday terms with a specific technical
meaning,resulting in an extremely frequent use of the term;these are words like
account,act,and adjust in Business,and address,arm,and assembly in Engineering.
The pattern of word use in Natural Science is quite different.Table 3.1 lists
a large number of specialized terms found exclusively in this discipline;most of
these words are highly technical in meaning and have no counterpart in everyday
usage.These are words like abscission,acastia,acetylation,achene,acyanogenic,etc.
Humanities and social sciences also have a large number of specialized words
that are restricted to these disciplines.However,as noted above,these words tend
to refer to concepts or entities that could easily be described in normal conversa-
tion with a fuller expression.For example:
altruism=caring about other people
ambivalence=having mixed feelings about something
amoral=doesn’t care about right or wrong
Detailed consideration of these specialized words shows that there are strikingly
different patterns of use across disciplines.Figures 3.5 and 3.6 show a general??University Language
Table 3.1 Specialized vocabulary beginning with the letter‘A’,broken down by academic
discipline
Business
Words found only(or primarily)in Business:
accredited,accrual,adjusted/ing,adversarial,affective/ity,amend,annualized,annuity,
apprentice,arbitration,audit,auditor,averse/ion
Words that are much more common in Business than in other disciplines:
accommodation,account,accounting,accumulated,accuracy,achievement,act(noun),action,
activate,activity,ad,adjust,administer,adopt,advantage,adverse,advertise,advertiser,affect,
affiliation,affirmative,agency,agent,agreement,allege,allocation,allowance,alternative,amend,
amendment,analysis,annual,approach,ask,aspect,assertion,assess/ment,asset,assurance,
attention,attractive/ness
Engineering
Words found only(or primarily)in Engineering:
absorber,adiabatic,aerodynamics,aerospace,algorithm,alloy,analyzer,annular,artifactual,
attenuate/ion,automated,axle
Words found only in Engineering and Science:
absorb,acetone,advection,airflow,algebraic,ammonia,ammonium,analog,angular,anisotropic,
annulus,anthracite,approximately,aqueous,aquifer,asbestos,aspherical,axis
Words that are much more common in Engineering than in other disciplines:
absolute,acceleration,access,accordingly,actual,address,alpha,ambiguous,amplify,amplitude,
analogous,applicability,application,apply,approximate/ion,arbitrary,architecture,arithmetic,
arm,array,arrow,assembly,assignment,associate
Natural Science
Words found only(or primarily)in Natural Science:
abscission,acastia,accretion,acetate,acetylation,achene,actin,acyanogenic,adnation,adsorb,
adsorption,advection,aeration,agaric,albumin,alder,algae,aliele,aliphatic,alkaline,
alleghaniensis,allotropic,allozyme,alluvial,alphape,aluminosilicates,amanita,amatoxin,
amphibole,anaerobic,andesite,anemophilous,anhydrite,anionic,anisotropic,anode,anoxic,
antiparallel,antiquark,antiviral,aperiodic,apetalous,aphyllophorale,apophysis,appendage,
arbuscules,archaebacteria,archean,ardente,arkose,armillaria,ascomycete,asepalous,asteroid,
asthenosphere,autochory,autoclastic,autodeliquescence,axil,axillary
Humanities and Social Science
Words found only(or primarily)in Humanities and/or Social Science:
altruism,ambivalence,amoral,anachronistic,anarchy,ancillary,anecdotal,angel,anger,anguish,
animosity,annexation,antagonism,anthology,antiquity,antithesis,antiwar,anxiety,apathy,
apostle,apostrophe,apprehend,apprehension,archaic,archetype,aristocracy,arousal,arrest,
arrogant,artifice,artisan,artistic,ascetic,ascribe,asocial,aspiration,assailant,assassin,assault,
assimilate,astonishment,astrology,astronomical,asylum,atheism,atonement,atrocity,attendant,
attentive,attribution,atypical,auspicious,austere,authenticity,authoritative,autobiography,
autocratic,autonomy,autopsy,avenge,avert,avid,awfulChapter 3.Vocabulary use in classroom teaching and textbooks??
similarity between Natural Science,Social Science,and Humanities:all three dis-
ciplines have a large number of different word types,including many words re-
stricted to a single discipline.However,this quantitative similarity corresponds to
very different patterns of word use in the disciplines.Natural Science uses a large
stock of technical terms for highly specialized reference:words that refer to entities,
characteristics,and concepts that are not readily discussed in everyday conversa-
tion.In contrast,Humanities and Social Sciences often offer new perspectives on
concepts and entities that are taken from our everyday experience.As a result,the
extensive stock of specialized words in these disciplines are often technical terms
to refer to these everyday experiences.
?.?Conclusion
This chapter has taken a different approach to vocabulary study than most pre-
vious studies of academic discourse,focusing on the diversity of vocabulary and
the frequency distributions of common and rare words across registers(rather
than the use of specific words or the identification of lists of the most common
words).This perspective exposed interesting contrasts between spoken and writ-
ten registers and across disciplines.Overall,classroom teaching was found to use
a relatively small set of different word types,but to rely heavily on a few of those
words,which therefore occur with extremely high frequencies.Textbooks,in con-
trast,were found to use a larger set of different word types(especially different
nouns),but none of those individual word types occur with extremely high fre-
quencies.Such distributions are consistent with the differences between the two
registers with respect to their planning and revising time,as well as the presence
of a face-to-face audience.
The comparison of vocabulary patterns across disciplines highlighted the re-
lationship between word use and subject matter.Business and engineering were
found to have less diversity in word choice,using words that are adapted from
everyday use but have specific technical meanings.Natural science has more di-
versity in vocabulary,reflecting the diversity of its sub-disciplines,and it has more
rarely occurring technical terms that have no everyday counterparts.The human-
ities and social sciences also evidence diverse word choice,consistent with their
diverse subject matter;many words in these disciplines could be summarized with
longer phrases in everyday language,as the disciplines address matters that are
more often part of our everyday experience.
Other perspectives on vocabulary use in academic language are of course
needed.For example,it would be useful to investigate the use of multi-word terms,
such as“hard income measure”in business or“truth value”in philosophy.Never-
theless,even the limited analyses possible with the relatively small sub-corpora in