Вы находитесь на странице: 1из 12

English for Specific Purposes 30 (2011) 164–175

Contents lists available at ScienceDirect

English for Specific Purposes


journal homepage: www.elsevier.com/locate/esp

The Discussion section as argument: The language used to prove


knowledge claims
Jean Parkinson ⇑
School of Linguistics and Applied Language Studies, Victoria University of Wellington, New Zealand

a r t i c l e i n f o a b s t r a c t

Article history: Writing the Discussion section of a laboratory report or dissertation is difficult for students
Available online 13 April 2011 to master. It involves complex causal, conditional and purposive argument; this argument
guides the reader from acceptance of the relatively uncontroversial data to acceptance of
Keywords: the writer’s knowledge claim. Students benefit therefore if they are assisted in acquiring
Discussion section the lexico-grammar commonly used in discussion of results. To explore the lexico-gram-
Research articles mar of Discussions, this article relies on two small corpora, one of physics research articles
Laboratory report
and the other of student physics laboratory reports. The article employs both a clause by
English for specific purposes
clause analysis and concordancing software to identify the key ways of expressing these
meanings. It finds the means employed in the student writing to be more congruent, more
emphatic and less closely argued than in the research article corpus, and suggests specific
grammatical resources which might form the subject of tasks from which students could
benefit.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Research Articles (hereafter RAs) are a much studied genre (Bazerman, 1988; Berkenkotter & Huckin, 1995; Swales, 1990).
Many features of the genre have been studied, including hedging (Hyland, 1996), reporting verbs (Thompson & Ye, 1991) and
the move structure of sections of RAs including Abstracts (Samraj, 2005), Introductions (Swales, 2004), Methods (Lim, 2006),
Results (Brett, 1994) and Discussions (Dudley-Evans, 1994; Holmes, 1997; Kamoksilapatham, 2005; Peacock, 2002; Swales,
1990). This article focuses on the Discussion section, the purpose of which is to state and explain findings, and demonstrate
to readers how the data collected prove the author’s knowledge claim.
This article examines the lexico-grammar used to argue and prove knowledge claims in the Discussion section. The
Discussion does this by embedding the writer’s data, methods, and references to literature in argument that adds up to
the Research Article’s claim. Prominent features in such argument are expression of causal, conditional and purposive mean-
ing, ways of asserting proof, and mental and verbal processes that either allow readers insight into the thought processes of
the author, or direct the readers’ thoughts. Before considering how these meanings are realised in my data, I will briefly
outline their description in the literature.
Simpson (2001, p. 596), who distinguishes cause, condition and purpose as three distinct categories, describes causal
argument as typically realised by means of the conjunctive adjuncts ‘so’, and ‘because’. Causal reasoning is structured as:

reason/cause (a), so effect (b).

Of relevance is Halliday’s concept (1993) of grammatical metaphor, which relates to meanings that are metaphorical with
respect to the grammar used to express them, rather than with respect to the meaning. Grammatical metaphor describes

⇑ Tel.: +64 44635233x8009.


E-mail address: jean.parkinson@vuw.ac.nz

0889-4906/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.esp.2011.03.001
J. Parkinson / English for Specific Purposes 30 (2011) 164–175 165

meanings expressed using grammatical means that are not the most congruent, that is the most expected or basic. Most con-
gruently, causal argument is realised through conjunctive adjuncts (Halliday, 1993, p. 66); less congruently cause can be rea-
lised verbally and nominally:

a happens so x happens most congruent


That a happens causes x to happen
Happening a is the cause of happening x least congruent

Causative verbs listed by Biber, Johansson, Leech, Conrad, and Finegan (1999, p. 417) include lead to, come from, result in,
contribute to, allow for and be required for. In addition, cause can be construed using the preposition due to:

Due to a, x happens.

Causal constructions are used in Discussions to explain the results or to explain why particular methods were used. Cause
and effect may appear in consecutive clauses (when a conjunction is used to express causal meaning) or within a single
clause. Either may come first in a sequence.
Conditional clauses express hypothetical meanings; if the conditional meaning is realised, the main clause event will also
be realised. Carter and McCarthy (2006, pp. 747–757) outline a range of ways that conditionality may be expressed. These
include subordinating conjunctions if, provided that, on condition that, assuming, in case of. Conditionals are typically struc-
tured (Simpson, 2001, p. 596) as:

IF condition p pertains, [then] it carries with it consequence q

This can also be realised as:

Condition p MEANS consequence q

Purposive arguments (Simpson, 2001, p. 596) suggest a course of action based on a targeted purpose/need, and are typ-
ically realised in the conjunctive adjuncts ‘in order to’ and ‘to’, but also in purpose circumstances, starting with ‘for’:

In order to achieve purpose/need x, do action y


To achieve purpose/need x, do action y
For purpose x, do action y

Of course, conjunctives are not the only way that purposive relations can be expressed. Heuttner, Vaughan and McDonald
(1987) include juxtaposition [I bought a book. I’m going to read it on the plane] causal conjunctive adjuncts [I bought a book so
that I could read it on the plane]; and infinitive clauses [I bought a book to read on the plane], as purpose clause types (examples
from Huettner, Vaughan, & McDonald, 1987, p. 207).

2. The data: research articles and student laboratory reports

I draw below on two corpora: physics RAs and undergraduate physics laboratory reports. I consider these together
because I regard laboratory reports as a pedagogical form of the experimental RA to the extent that they share the same orga-
nisation and both discuss experimental results. Being the major written form for undergraduate science students (Braine,
1989), and attached to the laboratory session, which is the key means by which students learn experimental methods as well
as the values of the discourse community, they teach young scientists the discourse features of science writing that
culminates in the Research Article.
The two genres have different purposes, and different audiences. The purpose of a student laboratory report is to
demonstrate knowledge and ability in performing process skills, and ability to relate their findings to theory. Its purpose
is pedagogical in the sense of teaching students to express experimental results in science and it is not expected to advance
new knowledge. The writer explains findings in the light of accepted knowledge, with deviations from accepted theory inter-
preted by both reader and writer as limitations in the writer’s experimental method rather than new findings. The writer
might be expected therefore to hedge less than a RA writer.
By contrast, new findings are essential in a RA, which seeks to gain acceptance of a new knowledge claim from the
research community. The writer must explain findings in the light of accepted knowledge, but also extend this knowledge.
RA writers therefore hedge more, because the information is new and not yet endorsed by the community (Myers, 1989).
Writers of student laboratory reports are comparatively inexperienced writers, who might, as I suggest below, be expected
to employ more congruent, ‘‘more spoken’’ means of expressing themselves, as well as using a more limited set of grammat-
ical resources to do so.
166 J. Parkinson / English for Specific Purposes 30 (2011) 164–175

3. Method

My data are two small untagged electronic corpora. The 204,000-word RA corpus comprises the Discussion sections of 30
articles selected from recent issues of 11 high impact Physics journals (ScienceWatch, 2008). (Publication dates were: 22 in
2010, 6 in 2009, 1 in 2008 and 1 in 2004.) Mean number of words per text was 6785. No distinction was made between na-
tive and non-native writers of English because the target community for a young scientist is an international one, most mem-
bers of which are not native-speakers, as a glance at the highly cited journals in the selected field (Physics) confirms.
The second 177,000-word corpus is 185 Discussion sections from undergraduate student Physics laboratory reports.
These were written from 2002–2007. The corpus was limited to one discipline, Physics, in order to minimise variation. All
writers were second language speakers of English at a South African university. All were first year students enrolled in a sci-
ence writing course and studying a range of science disciplines. The mean length of these Discussion sections was 960 words.
Two methods of analysis were used. The first was a clause by clause analysis of a selection of the texts, to identify mean-
ings being expressed, and to identify the grammatical resources used in expression of this meaning. Once these were iden-
tified, WordSmith 5.0 (Scott, 2008) was used to search the corpora for their use.

4. Results

This research relies on two kinds of analysis. A clause by clause analysis revealed the kinds of causal, purposive, and con-
ditional meanings being made and how these contribute to proving the writer’s claims of proof. Short extracts from this
qualitative data are discussed in section A. These reflect the analytical approach, and the insights it reveals about the causal,
condition and purposive resources used in the corpora. A concordancer was then used to search for these meanings to reveal
patterns in the data. These are discussed in Sections 4.1–4.6.

4.1. Cause, purpose, condition, and proof: clause by clause analysis

In this section I consider examples (Appendices A and B) of texts from each of my corpora. Appendix A is a section of one
of 11 texts from the RA corpus and Appendix B is a student Discussion section. These were analysed for statements indicating
cause and effect, condition and consequence, action and purpose as well as proof and the use of mental and verbal processes
in the argument. This section will point to features of these texts which I show below in my quantitative analysis to be trends
in the data. These trends are, firstly, the greater reliance by student writers (see Appendix B) on a smaller range of resources
for expressing causal condition and purposive meanings, particularly the use of conjunctions:

B10 measuring cylinder was rinsed with hot water


B11 so that the heat could not be lost.

By contrast the RAs (see example, Appendix A) express these meanings using a wider range of resources, relying less on
conjunctions and to a greater extent on adverbial phrases, nouns and verbs:

A20 Importantly, the behavior of the induced polarizations depends strongly on interparticle distance

The student writing is thus more ‘‘spoken’’, typically organised into short clauses joined with conjunctions. By contrast,
the RAs are highly nominalised with many embedded clauses and lexically dense; (the lexical density of Appendix A, calcu-
lated as lexical words per clause, is 8.9). The student text is, not unexpectedly, less lexically dense (lexical density of Appen-
dix B is 3.8).
A second trend which Appendix A illustrates is how ‘proof’ in RAs starts by displaying the data by use of relational pro-
cesses such as show and presents, used to refer the reader to graphs and figures, and mental processes such as see:

A10 Fig. 4 presents the extinction spectrum of a dimer of 36 nm silver spheres that are separated by 2 nm.

RA discussions build on this displayed data by using causal conditional and purposive argument (such as that in A12–24)
to explain to the reader what can be deduced from the data, thereby to lead the reader to stronger statements of proof such
as find in A32 below:

A31 We have examined the peak enhancements at other wavelengths,


A32 and we find that
A33 peak enhancement is associated with peak extinction . . .

Unlike Appendix A, Appendix B does not refer the reader explicitly to elements of the data, expecting the reader to do this
work unaided. Apparently the reader is regarded as one who knows what the ‘‘correct answer’’ is and does not need the kinds
of persuasion represented in starting with the proof reflected in the data and gradually building up to strong claims of proof.
As I show below, this was a trend in the data; student writers were less likely to make the ‘weak’ claims of proof involving
J. Parkinson / English for Specific Purposes 30 (2011) 164–175 167

pointing the reader to the data; they were more likely than were RA writers to make ‘strong’ claims of proof, involving claim-
ing that the data proves or confirms some physical principle, using less tentative language than that of the RAs. This can be
seen in clauses B24–B29, which are unhedged and definite.
I turn now to the ways that meanings used in argument in Discussions are expressed in my data; I consider below how
writers in my two corpora express cause, conditional meaning, purpose, proof and use mental and verbal processes to con-
vince their readers of the claims they make about their experimental data.

4.2. Expression of cause

As discussed in Section 1 above, cause can be realised between clauses using conjunctions, or within the clause, nomi-
nally, verbally or via the preposition due to.

4.2.1. Cause between clauses: conjunctives


As Table 1 shows, conjunctions were not prominent in the RA corpus. The most common however was Thus:
(1) The drag coefficient can be reduced. . . but this leads to faster rotation, and thus requires greater time resolution. (RA
B&B).

Because was the most frequent means of expressing causal meaning in my student corpus, with other frequently used
conjunctions being Therefore and Since.
(2) Time is also a factor because if you take more time to do your experiment the hot water will cool down (Student lab-
oratory report).

4.2.2. Cause within the clause


Conjunctives are just one way that causal relations can be expressed. Other, less congruent, ways of expressing these
meanings include the preposition due to and causal nouns and verbs:
Adverbial phrases:

(3) This may be due to the amorphous nature of the as-deposited films. (RA FN&K)

Table 1
Causal meaning in the two corpora.

Causal meaningsb Student writing Research articles Significant at 95% levela


Number % Number %
Adverbial Due to 52 4 30 11 Yes
Associated with 0 11 4 Yes
Adjectives Caused 2 3 1 No
Resulting 5 1 No
Conjunctions As a result 13 1 7 2 Yes
Because 429 32 17 6 Yes
Since 71 5 12 4 No
So 36 3 6 Yes
Therefore 84 6 17 6 No
Thus 31 2 33 12 Yes
Subtotal 664 50 103 37
Noun Cause 3 1 No
Consequence 5 4 1 No
Influence 8 1 10 4 Yes
Reason 35 3 3 1 No
Result 307 23 53 19 No
Subtotal 358 27 71 25
Verbs Arise 1 4 1 Yes
Cause 137 10 9 3 Yes
Contribute 21 2 4 1 No
Influence 5 2 1 No
Lead/led 23 2 25 9 Yes
Make 25 2 9 3 No
Result 32 3 10 4 No
Subtotal 244 18 63 22
Total expressions 1325 282
a
An alpha level of .05 was chosen as the significance level. A significant difference reflects whether the two corpora differ in their proportion of use of a
particular resource for expressing cause as a proportion of all resources used in each corpus.
b
Concordance data was sorted to exclude uses of these words that did not fall into these categories.
168 J. Parkinson / English for Specific Purposes 30 (2011) 164–175

Causal nouns (result, influence, consequence, contribution, cause, reason):

(4) The reason this was so is because we started with a high volt ..(Student report).

Causal verbs include lead/led, cause, result, make, arise, contribute, and influence.

(5) This mechanism stiffens the CDW and leads to an increase of the longitudinal phason velocity [15]. (RA R&D)
(6) The height the swing reach depends on the potential energy, which is said to be caused by the length of the chains. . .
(Student report)
(7) The single-stranded DNA form is more favorable than when in solution, which would result in a decrease in DNA melt-
ing temperature. (RA B,H&B)

Realisation of cause within a clause may make the reasoning less arguable than if it is in two consecutive clauses, as is
more common in student writing. The cause or effect may be found too in a rank-shifted clause, such as in (7) above, which
also functions to make the reasoning less arguable.
Conjunctions are the most congruent expression of causal meaning, so we would expect inexperienced writers to use
them more frequently than RA authors. Table 1 reflects less use of a wider range of causal conjunctives in RAs, and higher
conjunctive usage in the student writing, particularly of because. As much as 50% of the causal meaning in the student corpus
is expressed by means of conjunctions, with because representing a third of all causal meanings in the corpus; in the RA cor-
pus 37% of causal meaning is expressed by means of a wider range of conjunctions, the most prominent being thus. Biber
(2006) found causative subordinate clauses (starting with because) to be more common in spoken than written university
registers; they were most common in study group discourse where students use these clauses to ‘‘provide supporting argu-
ments or justifications for proposed answers or explanations’’ (Biber, 2006, p. 79). Familiarity with study group discourse
might be one of the reasons for the ‘‘less written’’ register in student lab report Discussions. The noun Result is prominent
in both corpora (23% of causal meanings in the student corpus and 19% for the RA corpus). The verb Cause makes up 10%
of causal meanings in the student writing, while in the RAs the most prominent verb is lead/led (9%).
Table 1 also shows that student writing uses more limited expression of causal meaning, with just six words making up
80% of the causal meanings counted in the corpus (because, result (noun), cause (verb) therefore, since, and due to). In the RA
corpus ten words make up 81% of the meanings counted in the corpus (result (noun), thus, lead/led, because, so, therefore,
since, associated with, influence (noun) and result (verb)). Table 1 indicates that use of most frequently used words expressing
causal meaning is statistically different in the two corpora; however comparison of use of individual words is less important
than the trend Table 1 displays that the student corpus is more likely to construe meaning using conjunctions, and that over-
all a smaller range of words is used to express causal meaning in the student corpus than in the RAs.
A look at Appendix A illustrates how authors use causal meanings to build up to and support knowledge claims. In clauses
A20–A22 (Fig. 1, from Appendix A), readers are lead to see that the data depend on or are influenced by particular factors;
readers are lead from the ‘weak proof’ that is evident when they are asked to look at what the results in Fig. 4 presents
and shows (clauses A10 and A11) to clauses A31 and A32 where the author make the stronger expression of proof repre-
sented in we find that. . . A number of such stronger expressions of proof, over longer sections of the Discussion, collaborate
to add up to the author’s knowledge claim.

4.3. Expression of conditional meaning

Table 2 indicates that resources for expressing conditional meaning in the data include the conditional conjunctions if,
once, when, after; adverbial phrases which imply contingency, and verbs indicating conditional meaning. (Concordance data
was sorted to exclude instances when these words did not indicate conditional meaning.)

4.3.1. Conditional conjunctions


Examples are:

(8) If the released DNA rehybridizes. . ., the amount of fluorescent DNA in solution will be significantly higher. . .. (RA
B,H&B)
(9) When the turns around the iron nail were constant . . .only two pins were attracted. (Student report)

4.3.2. Adverbial phrases that express contingency


Examples in this category are:

(10) Under the experimental conditions described above (RA B,H&B)


(11) If I were to do this experiment again, for current vary and turns constant, I would start from the lower volts. (Student
laboratory report).
(12) For the spheroid and rod, the long wavelength resonance red shifts with increasing aspect ratio, and at the resonance
wavelengths, the E-field enhancements increase with increasing aspect ratio (RA H&S)
J. Parkinson / English for Specific Purposes 30 (2011) 164–175 169

A10. Figure 4 presents the extinction spectrum of a dimer of 36 nm silver spheres


that are separated by 2 nm.
A11. Included are three spectra: One for polarization parallel to the interparticle axis
that shows a strong peak at 520 nm and a shoulder peak at 430 nm, one for
perpendicular polarization that shows a peak at 410 nm, and one that corresponds
to an average over the three possible polarizations.

A20. Importantly, the behavior of the induced polarizations depends strongly on


interparticle distance,
A21. with dipoles in the interfacial region between the spheres being strongly
influenced by interparticle coupling when the particles are close together.
A22. As a result, the extinction spectra of dimers are very sensitive to separation
distance, with the dipole resonance red shifting, and the quadrupole resonance
becoming more intense as the particles approach.

A31. We have examined the peak enhancements at other wavelengths,


A32. and we find that
A23. peak enhancement is associated with peak extinction for all the structures that
we have considered.
Fig. 1. How causal meaning leads readers from the data to the knowledge claim.

Table 2
Expression of conditional meaning in the two corpora.a

Student corpus Research articles Significant at 95% levelb


Number % Number %
Conjunction
If x. . .(then) y 146 28 9 4 Yes
Once x. . .(then) y 3 1 3 1 No
When x. . .y 340 66 53 26 Yes
After x. . .y 2 4 2 Yes
Subtotal 491 95 69 34
Adverbial
At x. . .y 10 2 69 34 Yes
For x. . .y 1 18 9 Yes
With x. . .y 10 2 25 12 Yes
Under condition x. . .y 2 15 7 Yes
Subtotal 23 4 127 62
Verbs
Depending on x. . .y 3 1 3 1 No
Using x. . .y 0 5 2 Yes
Total 517 204
a
Concordance data was sorted to exclude use of these words that did not express proof and to distinguish the two meanings of show.
b
An alpha level of .05 was chosen as the significance level.

4.3.3. Verbs indicating conditional meaning

(13) Depending on the method used (RA F,M&T)


(14) about 8.5-fold enhancement can be produced using the target with porosity 95%. (RA F,M&T)

Table 2 indicates that use of the conjunctions if and when is greater in the student writing than in the RAs, and that these
are the main resources used in the student writing to express conditional meaning. The RA corpus used a wider range of
resources to express conditional meaning, in particular the conjunction when, and adverbial phrases with at, for, with, and
under. Biber (2006, p. 77) found that conditional subordinate clauses (starting with if), like causative subordinate clauses,
170 J. Parkinson / English for Specific Purposes 30 (2011) 164–175

A12. To assign the resonances,


A13. we examined the induced polarization in each DDA element at the different
wavelengths.

A15. If we were to expand the induced polarization in vector spherical harmonics


about the midpoint of the spheres,
A16. the 520 nm polarizations would be predominantly dipolar in character
A17. while those at 430 nm would be quadrupolar,
A18. so we will use the terms "dipole plasmon" and "quadrupole plasmon"
A19. to label these features.
Fig. 2. Conditional meaning functions to explain how experimental work was done.

are more common in spoken than written university registers. They were particularly prominent in office hours, in which
conditionals support explanations in individual teaching (2006, p. 78). Thus it could be argued that students are most used
to these more ‘‘spoken’’ means of expressing reasoning, and that this might be a reason for the ‘‘less written’’ register used in
laboratory report Discussions. Students would benefit from exposure to the range of means for expressing conditional mean-
ing, perhaps through looking at concordance data.
Conditional meaning (e.g. clauses A15–A17, Fig. 2) works with causal and purposive meaning to take the reader from
the ‘weak’ claims of proof of the bare data to the ‘strong’ claim of proof of the authors’ claim. Specifically, it functions to
show the conditions under which the data behaves the way it does or, in Fig. 2, why the experimental work was done as
it was, and how a particular method was necessary to fulfil the authors’ purpose (expressed A12–A13 and A18–A19).

4.4. Expression of purpose

Authors use purposive meaning to express overall purpose in their research (see A3–4, Appendix A), and, more fre-
quently, to explain their purpose in performing certain actions within the research (15–18 below). Thus information
about aims and methods crucial to the readers’ understanding of the research may be expressed in statements about pur-
pose and action.
In my data, purpose was expressed using the conjunctive adjuncts in order to, to, so that, and so as. (Concordance data was
sorted to exclude instances not related to purpose.) As Table 3 shows, to was the most common expression of purpose in both
corpora. Example 16, like A19 in Fig. 2, clearly expresses the author’s purpose with respect to use of their experimental
method.

(15) We have used nonlinear methods to assess the relative importance of the two oscillatory systems in allowing synchro-
nization to be achieved (RA SS&M).

In addition, purpose was expressed by means of phrases of the form by + gerund:

(16) Kondo explained this unusual behaviour of S by considering the fourth-order perturbation.(RA TT&S)
(17) the SMF was reduced by lowering the external Naþ concentration to 0.1 mM (RA B&B)

Finally, rarely, the purpose is found in an adverbial phrase which follows the action.

(18) Both of these structures may be applied to fuel cell applications . . . for the conversion of chemical to electrical energy
(RA NN&K).

Table 3
Expression of purpose.

Student corpus Research articles


Conjunctive adjuncts
In order to 12 7
to 155 44
So that 21 7
So as 6 0
Total instances 194 58
J. Parkinson / English for Specific Purposes 30 (2011) 164–175 171

4.5. Expression of proof

Proving the author’s knowledge claim is the purpose of the RA, and in particular of the Discussion. Yet, as has been dis-
cussed in the literature (e.g. Hyland, 1996, 1998, 2005; Myers, 1989), writers must use a range of persuasive means to do
this. An important one for the present study involves starting from the undeniable proof of the data (which I have called
‘weak’ claims of proof), to advance gradually through explanations involving cause, condition and purpose to the less obvious
proof (‘strong’ claims of proof) of the author’s knowledge claim. The author begins by pointing to (displaying/showing/present-
ing/giving) and explaining the data in such a way that demonstrates to the reader that it amounts to (shows/demonstrates/
means/confirms) and proves the knowledge claim. Key in this process is a range of mental processes through which the reader
is asked to see/observe the data, or is given insight into the perspective from which the author considered the data. The data
itself is also given voice and suggests certain interpretations.
Verbs such as show, indicate, reveal and mean, which express proof in my corpora, make ‘weak’ and ‘strong’ claims of
proof. The weak claim of proof includes words with a meaning approximating ‘show’. Examples include show, present,
and give. These words are usually used to refer the reader to data that has been displayed in graphs. By contrast with the
stronger claims of proof, their evidence is plain to any reader who merely looks at the graphs.
Most commonly in the RA corpus show means ‘display’, a weaker more literal claim of proof:

(19) . . . the derivative plots are shown as insets. (RA BH&B)


(20) Included are three spectra: One for polarization parallel to the interparticle axis that shows a strong peak at 520 nm . . .
(RA H&S).

The verb present appears to have the same meaning as show:

(21) . . . the thermal conductivity enhancements present a peak value of 27.5% . . . (RA X&C)

Give is used with a slightly stronger meaning, and usually points to the link between displayed data and its mathematical
expression:

(22) The flux is thus given by 4 U(s) = q[y(s)](dy/ds) (RA CJ&K)

The verbs constituting weak claims of proof, show (meaning display), give and present make up most of the meanings
concerning proof that were counted in the RA corpus. Only once the argument has been made using such weaker indica-
tions of proof, do RA authors use the stronger indications of proof such as show (meaning prove), confirm, demonstrate, and
mean. Weaker indications of proof are far more frequently used in the RA corpus than the stronger words show (meaning
prove), confirm, demonstrate and reveal, which make up only around 15% of the meanings concerning proof in the corpus
(Table 4).
The following are examples of where show is used to make a strong claim of proof. Interestingly in example 23, this strong
claim actually refers to a result already published in the literature (referenced as [9]). Having been published means that it
has some acceptance by the research community, and has gone some way to being accepted as fact (Myers, 1989) and thus
can be stated more baldly.

(23) These results are consistent with previous ones [9] . . .which show that the Peierls distortion leads to Mo atoms dis-
placements . . . (RA R&D)
(24) The E-field contours show that the maximum enhancement for the dipole resonances occurs at the particle tips. (RA
H&S)

Indicate, like show, also seems to have this strong meaning of ‘prove’:

(25) A small value of the geometric standard deviation indicates a narrow particle size distribution. (RA X&C)

Table 4 suggests that the strong claim of proof is used less sparingly in the student writing, which employs a high pro-
portion of verbs meaning ‘prove’: show (meaning ‘prove’), prove, and mean. These are strong words, lacking the tentativeness
that is more apparent in the RA corpus. Together with similar less frequently used words such as confirm, demonstrate and
indicate, these strong meanings make up 75% of the words claiming proof in the student corpus.
The near absence of prove and proof in the RA corpus is an example of the tentative and subtle way that RA authors pro-
vide proof for what they say. To say that they have proved something is too blatant and too lacking in caution. By contrast,
prove accounted for 15% of the much less tentative student use of language indicating proof. The use of prove in the following
extract from a student report makes for a strong claim:

(26) The results that were also obtained from the swing with shorter ropes were also interesting, because swing proved
that it does not depend on force applied. (Student report)
172 J. Parkinson / English for Specific Purposes 30 (2011) 164–175

Table 4
Expression of proof in the two corpora.a

Student writing Research articles Significant at 95% level


Number % Number %
Noun
Proof 3 0
Material processes
Give 2 24 7 Yes
Present 0 20 6 Yes
Subtotal 2 44 14
Identifying processes
Demonstrate 3 15 5 Yes
Confirm 20 6 2
Indicate 12 3 18 6 No
Mean 69 19 3 Yes
Prove 53 15 1 Yes
Reveal 3 7 2
Show (prove) 109 31 26 8 Yes
Show (display) 63 18 107 33 Yes
Subtotal 332 92 178 55
Mental processes
Observe 6 2 27 8 Yes
See 10 3 53 16 Yes
Subtotal 16 6 80 25
Verbal process
Suggest 3 22 7 Yes
Total counted 356 325
a
Concordance data was sorted to exclude the use of these words that did not express proof and to distinguish the two meanings of show.

Claims of proof are also found in the mental and verbal processes of observing, seeing, expecting, suggesting and explaining.
Most frequently, the mental and verbal processes I list here function not, as we might expect, to afford insight into the rea-
soning processes of the author or reader, but rather to suggest that the reader accept the proof of their own (or, in example
29, Table 5, the author’s) perceptive processes. As proof of this I provide examples from both corpora in Table 5. An exception
in Table 5 is the student use of suggest (example 27, Table 5), which, although rare in the corpus, clearly functions as an inter-
textual voice (that of the textbook) rather than to provide proof. This is one example of the very different way that certain
words are used by expert and student writers, and the probable value of exposing students to expert means of expression.
The reader is guided to see (examples 31 and 32) in the data the proof that the author is demonstrating in the argument.
In example 30 observe is clearly an equivalent of show, while in example 29 proof is provided in the form of the writer’s at-
tested perception: we observed. Use of suggest (example 28) is an interesting means of allowing the data, rather than the
author, to draw certain conclusions. Because the results apparently speak for themselves, this allows the data to make a
slightly stronger claim that mere showing would do:

4.6. The reasoning processes of reader and writer

Reference to the writer’s and reader’s mental and verbal processes (compared in Table 6) also play a role in argument, and
their use is highly persuasive in intent. They afford the reader an insight into the writer’s thought processes, or serve to direct
the reader’s thought processes. As found in analysis of other resources in this article, the RA writers employ a wider range of
resources (including observed/observable, expect, explain, know and consider) than do the student writers (who rely heavily on

Table 5
Mental and verbal processes as expression of proof.

Student corpus RA corpus


Suggest (27) We will be able to notice if the number of pins (28) The above results suggest other dominant factors (RA X&C)
double as the book suggest
Observe (29) We observed that 24 pins were picked up with 7 (30) They performed a fluctuation analysis which observed a quadratic
loops relationship (RA F&R)
See (31) As you can see that the hot water has a bigger (32) In Fig. 4a it can be seen that annealing has led to an increase in the
amount than cold water refractive index of the film with the values of 2.4 and 3.3 at 350 nm . . .. (RA
FN&K).
Confirm (33) Therefore the hypothesis is correct and confirms (34) Further experiments confirmed that each step uses one molecule (RA B&B)
that as the current and number of coils increased. . .
J. Parkinson / English for Specific Purposes 30 (2011) 164–175 173

Table 6
Reasoning processes.

Student writing Research articles Significant at 95% level


Number % Number %
Adjectives Considered 0 4 3 No
Known 1 4 3 No
Observed/observable 0 14 10 Yes
Predicted 0 2
Subtotal 1 24 18
Mental processes Ascribe 0 3 2
Conclude 20 12 3 2 Yes
Expect 94 54 17 13 Yes
Explain 10 6 15 11 No
Know 9 5 12 9 No
Predict 9 5 2 No
Think/thought 11 6 2 No
Consider 6 4 30 22 Yes
Subtotal 159 92 84 62
Verbal processes Describe 2 10 7 Yes
Nominalised mental processes Consideration 3 2 3 No
Description 0 1
Expectation 2 0
Explanation 1 4 3 No
Observation 2 6 4 No
Prediction 3 2 1 No
Indication 0 2
Subtotal 11 6 18 13
Total counted 173 136

expect and conclude). As indicated in Table 6, there is a significant difference (95% level) in use of most of the frequently used
words in the corpora.
Consider affords the readers insight into the authors’ mental processes (example 35), and could be viewed as a way of
taking the reader along with the author’s argument. Consider is also used to direct the reader’s attention to data (example
37) and in example 36 as a directive to the reader to accept the author’s claim:

(35) First, we consider the nature of LRO below T0. . .(RA TT&S)
(36) . . . cold electroweak baryogenesis should be considered a serious candidate scenario for. . . (RA TH&K)
(37) . . . we have investigated the effect of Gd on the band structure by considering 2 values of m . . . in Eq. (3) . . . (RA FN&K)

With regard to the high incidence of expect in the student corpus, students are likely to have been influenced by the
guidelines they were given for laboratory report writing (see Fig. 3).
This is likely to have increased the use of the word expect, and demonstrates the reliance of students on guidance of this
sort and the need for such guidance to reflect the target register.
Directing of the reader to observe the data and interpret it as proof are again reflected in nominalised mental processes
such as observation and explanation and nouns such as proof and indication:

(38) These observations suggest that the surface of the magnetite particles was coated efficiently with PMAA (RA TO&V).

Table 6 indicates rather different use of mental processes and other means of affording insight into reasoning processes in
the corpora. It indicates how carefully argued RAs are, with data being introduced to be considered, and what is observed/ob-
servable/known being carefully built upon to support the authors’ claim. By contrast focus in the student corpus is on what
the author expected or can conclude. This suggests the value of sharing concordance examples with students.

Discussion
First of all, most importantly, explain the data. In addition, consider how your data
compares with what you expected from theory? How does it differ? Try to explain
any differences from what you expected. What factors could have contributed to any
unexpected results?

Fig. 3. Extract from report structure guidelines.


174 J. Parkinson / English for Specific Purposes 30 (2011) 164–175

5. Discussion

I have argued above that the Discussion section provides an argument that leads the reader from the proof of the data
(whose meaning is open to inspection in graphs and tables and is, if the author’s method is accepted, uncontroversial) to
the proof of the claim. The claim is not open to inspection merely by inspecting the data and requires careful argument con-
cerning the cause of the results, and the conditions required for the results. This journey from data to claim is thus achieved
via explanations about the cause of elements of the data, about the purpose of performing the experimental work in the way
that it was performed, and about the conditions needed for the experiment to function.
The Discussion section draws on a wide range of lexico-grammatical resources to achieve these meanings. A difference
between the two corpora in expressing these meanings is that the student writing is more likely to express meanings more
congruently using ‘more spoken’ language. For example, there is relatively high use of the conjunctions if and when to ex-
press conditional meaning in the student writing, but the RA corpus relies more heavily on adverbial phrases. Similarly a
high proportion of causal meaning in the student corpus is expressed using the single conjunction because, while the RAs,
while avoiding the less tentative constructions such as the verb cause, use a wider range of conjunctions as well as a wider
range of nouns and verbs. This reflects a general trend in which RAs employ a wider range of resources to express these
meanings.
There was also a tendency for RAs to be more cautious and tentative. This difference is evident in expression of proof in
the corpora. In RAs show more frequently means ‘display’, while in the student writing it more frequently means ‘prove’.
There is high incidence too of strong words such as prove and mean in the student corpus and little or no use in the RAs.
In addition, the RAs have a high incidence of mental processes to suggest proof including invitations to observe/see the data,
insights into the way the authors considered the data, and the interesting use of suggest with the data itself as the sayer/sub-
ject, thus allowing the data to speak.

6. Conclusion

The article finds that students’ expression of cause, condition and purpose is more congruent than that in the RAs. Stu-
dents are more likely to employ conjunctions, the ‘‘more spoken’’ means of expressing causal and conditional meaning than
RAs are. With respect to claims of proof, the student writing in this study is more emphatic and less tentative. It makes more
strong claims of proof and fewer weak claims of proof than the RAs. By not using words like ‘show’ to point to the evidence of
the data, students rely on the reader to do more of the work of interpreting their data. They skimp on weak claims that show
the data. They are less likely than RA writers to employ mental processes asking the reader to see the data, or to share the
perspective from which they considered the data. Instead they jump more quickly to making strong claims of proof.
Another finding of this article is that, as might be expected, students used a more restricted range of lexico-grammatical
resources. Although space does not allow me to illustrate this, students could benefit from tasks based on concordance data
(c.f. Mudraya, 2006), or from being asked to notice and identify (c.f. Hyland, 2008) the lexico-grammatical means by which
causal meaning or claims of proof are being realised in suitable sections of text (perhaps something similar to, but more
simple than, the text in Appendix A). These findings of greater congruence, less tentativeness and a more restricted range
of lexico-grammatical resources all point to the need to widen teaching to include a focus on other lexico-grammatical
means, and the need to employ more closely argued reasoning in leading the reader from the proof represented by the data
to making the final claim of proof.

Acknowledgements

I would like to thank Dr. Elaine Vine and two anonymous reviewers for valuable feedback on an earlier version of this
article.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.esp.2011.03.001.

References

Bazerman, C. (1988). Shaping written knowledge. Madison: University of Wisconsin Press.


Berkenkotter, C., & Huckin, T. N. (1995). Genre knowledge in disciplinary communication: Cognition, culture, power. Hillsdale, NJ: Lawrence Erlbaum.
Biber, D. (2006). University language: A corpus-based study of spoken and written registers. Amsterdam: John Benjamins.
Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman grammar of spoken and written English. Harlow: Pearson Education.
Braine, G. (1989). Writing in science and technology: An analysis of assignments from ten undergraduate courses. English for Specific Purposes, 8(1), 3–15.
Brett, P. (1994). A genre analysis of the results section of sociology articles. English for Specific Purposes, 13(1), 47–59.
Carter, R., & McCarthy, M. (2006). Cambridge grammar of English. Cambridge: Cambridge University Press.
Dudley-Evans, T. (1994). Genre analysis: An approach to text analysis for ESP. In M. Coulthard (Ed.), Advances in written text analysis (pp. 219–228). London:
Routledge.
Halliday, M. A. K. (1993). On the language of physical science. In M. A. K. Halliday & J. R. Martin (Eds.), Writing science (pp. 54–68). London: The Falmer Press.
J. Parkinson / English for Specific Purposes 30 (2011) 164–175 175

Holmes, R. (1997). Genre analysis, and the social sciences: An investigation of the structure of RA discussion sections in three disciplines. English for Specific
Purposes, 16(4), 321–337.
Huettner, A.K., Vaughan, M.M., & McDonald, D.D. (1987). Constraints on the generation of adjunct clauses. In: Proceedings of the 25th annual meeting on
association computational linguistics. Stanford: CA, pp. 207–214.
Hyland, K. (1996). Writing without conviction? Hedging in scientific RAs. Applied Linguistics, 17(4), 433–454.
Hyland, K. (1998). Hedging in scientific RAs. Amsterdam: John Benjamins.
Hyland, K. (2005). Metadiscourse. London: Continuum.
Hyland, K. (2008). As can be seen: Lexical bundles and disciplinary variation. English for Specific Purposes, 27, 4–21.
Kamoksilapatham, B. (2005). Rhetorical structure of biochemistry RAs. English for Specific Purposes, 24, 269–292.
Lim, J. M. H. (2006). Method sections of management RAs: A pedagogically motivated qualitative study. English for Specific Purposes, 25(3), 282–309.
Mudraya, O. (2006). Engineering English: A lexical frequency instructional model. English for Specific Purposes, 25, 235–256.
Myers, G. (1989). The pragmatics of politeness in scientific articles. Applied Linguistics, 10, 1–35.
Peacock, M. (2002). Communicative moves in the discussion section of RAs. System, 30, 479–497.
Samraj, B. (2005). An exploration of a genre set: RA abstracts and introductions in two disciplines. English for Specific Purposes, 24(2), 141–156.
ScienceWatch (2008). Sci-bytes journals ranked by impact: Physics. <http://www.in-cites.com/research/2001/february_26_2001-2.html> (accessed
01.03.10).
Scott, M. (2008). WordSmith tools version 5. Liverpool: Lexical Analysis Software.
Simpson, P. (2001). ‘Reason’ and ‘tickle’ as pragmatic constructs in the discourse of advertising. Journal of Pragmatics, 33, 589–607.
Swales, J. M. (1990). Genre analysis: English in academic and research settings. Cambridge: Cambridge University Press.
Swales, J. M. (2004). Research genres: Explorations and applications. Cambridge: Cambridge University Press.
Thompson, G., & Ye, Y. (1991). Evaluation in the reporting verbs used in academic papers. Applied Linguistics, 12(4), 365–382.

Source text

Hao, E., & Schatz, G. C. (2004). Electromagnetic fields around silver nanoparticles and dimmers. Journal of Chemical Physics, 120(1), 357–566.

Jean Parkinson lectures Applied Linguistics/TESOL at the University of Victoria in Wellington. Her main research interests are the discourse features of
science genres and academic literacy acquisition, particularly literacy acquisition in the areas of science and applied science.

Вам также может понравиться