You are on page 1of 38

Semantico-syntactic environments of the

verbs show and demonstrate and Spanish


mostrar and demostrar in a bilingual corpus
of medical research articles

Ian A. Williams
Universidad de Cantabria (Spain)

The verbs show and demonstrate, and their potential Spanish counterparts
mostrar and demostrar, are frequent lexical verbs appearing in various settings in
medical research articles (RAs). This study analyses the contextual environments
of these verbs in an extensive corpus of medical RAs, composed of three subcor-
pora: English source texts, their Spanish translations, and comparable Spanish
native language texts. The verbs are analysed in terms of syntax (active and
passive) and the semantics of the main associated noun: Characteristics, Au-
thors, Evidence, Techniques, and Metatextual. The study uses quantitative and
qualitative methods in a three-way analysis: intralinguistic analysis compares the
environments for the verb pairs in English and Spanish; interlinguistic analysis
assesses similarities and differences in the environments between the two native
language subcorpora; and comparison of source and target texts provides insight
into translation behaviour. The implications for translation are discussed in
terms of context, collocation and appropriateness of discourse style.

Keywords: specialised corpora, comparable corpora, parallel corpora,


quantitative analysis, contrastive analysis, medical translation, Spanish-English
translation, research articles

1. Introduction

Comparable texts, that is similar native language (NL) texts in two or more lan-
guages, are one of the translators most useful tools (Neubert & Shreve 1992; Wil-
liams 1996a). Linguistic analysis based on comparable corpora provides valuable
data on frequency of use and insight into contextual and combinatory patterns of

International Journal of Corpus Linguistics 13:1 (2008), 3874. doi 10.1075/ijcl.13.1.04wil


issn 13846655/e-issn 15699811  John Benjamins Publishing Company
Show, demonstrate, mostrar and demostrar in medical articles 39

lexical items (Laviosa 2002; Tognini-Bonelli 2001). Parallel texts, i.e. pairs of source
language (SL) and target language (TL) texts, give information on actual transla-
tion behaviour. Corpora that include both a comparable and a parallel component,
therefore, allow the analyst to compare this behaviour with NL text usage. This
type of contrastive analysis is particularly fruitful when dealing with polysemous
lexical items involved in one-to-many or many-to-many relations, and can be ex-
pected to reveal the environments in which correspondences are consistent and
contexts in which deviations are seen to occur (Williams 2007, forthcoming).
In medical research articles, the verb show is one of the most frequent lexical
verbs (Williams 1996b). It is polysemous and appears in connection with a number
of rhetorical functions including the signalling particularly in the Results sec-
tion of non-textual material in what is called a metatextual pointer (Brett 1994;
Williams 1999), and in Dudley-Evans (1994) rhetorical model of the Discussion
section, for reference to previous research (RPR), statement of results (SOR), and
staking of claims. It is closely related to the less frequent verb demonstrate, with
which it shares a number of functions.
In a wider context of academic prose, Biber et al. (1999) categorise show as
an activity verb and, in a more specialised role, as a communication verb. They
note that the exceptional versatility in the syntactic patterning of this verb is reg-
ister-dependent; none of the patterns involving the indirect object was attested
in academic prose, where the outstanding patterns were S + V + DO and S + V +
that-clause. In this register, show is associated with a high percentage of inanimate
subjects, these entities being in some way instrumental to the meaning of the verb.
Show was also found to participate in 4-word and 5-word bundles: are shown in
table and is shown in figure/fig. in relation to tabular and graphic displays; and
studies have shown that and has been shown that in the reporting of previous re-
search. These data lend support to the significance of show in medical discourse.
However, there are no studies comparing show with demonstrate, nor from the
translation viewpoint are there contrastive data relating these two English verbs
with their apparent Spanish counterparts mostrar and demostrar.
This paper examines the relation between show and demonstrate and mostrar
and demostrar in an extensive bilingual corpus of medical research articles. The pa-
per is organised as follows. Section2 describes the design of the specialised corpus
of medical research articles and establishes the semantico-syntactic framework
of environments for interlinguistic and intralinguistic analysis. Section3 presents
the quantitative results and a discussion of the differences observed in the qualita-
tive contextual analysis. Section4 illustrates the actual translation behaviour re-
vealed in the comparison of the SL and TL texts and, finally, recommendations are
provided based on the data from the Spanish NL texts.
40 Ian A. Williams

2. Study design

2.1 The corpus

The corpus used in this study contains approximately half a million words and
consists of 192 complete research articles (RAs) with the typical Introduction,
Methods, Results and Discussion (IMRAD) format, but excluding the Abstract,
which is considered a separate genre. The corpus is divided into three subcor-
pora: a subcorpus of 64 Spanish texts translated from English (STLC), composed
of eight randomly selected RAs from each of the eight medical journals with Span-
ish editions in 1993; a subcorpus of the 64 English SL texts (ESLC); and a com-
parable subcorpus of 64 Spanish NL texts (SNLC) in sets of eight RAs randomly
selected from Spanish journals to cover the same specialities as ESLC and STLC.
The specialities covered are general medicine (two journals in each subcorpus),
cardiology, dermatology, gynaecology and obstetrics, ophthalmology, paediatrics,
and surgery (one journal in each case for ESLC and STLC, and eight RAs from one
or more journals for SNLC).

Table1. Characteristics of the three subcorpora of medical research articles


Characteristic ESLC SNLC STLC
Number of articles 64 64 64
Total number of words 157,650 140,250 185,000
Mean number of words per article 2,463 2,191 2,890
Total number of sentences 6,740 5,033 6,726
Mean number of words per sentence SD 23.2 3.1 28.0 4.9 27.3 4.0
Lexical to grammatical item ratio 55 : 45 44 : 56 44 : 56
Note: SD = standard deviation

Table1 gives some characteristics of the corpus. SNLC is somewhat smaller than
ESLC and STLC. The TL texts show a 17.4% increase over their SL texts, the differ-
ence being almost entirely due to grammatical items such as prepositions and ar-
ticles, which are added as English complex nominal groups with premodification
are transferred into Spanish with postmodification. The mean number of words
per sentence, calculated as the average of the mean words per sentence of each
individual article, was not significantly different in SNLC and STLC, but sentences
in ESLC were significantly shorter, again attributable to the different ratio of full
lexical to grammatical items. The data support the comparability of the subcor-
pora for the verbs under study, since these almost always appear in finite form and
are, therefore, a function of the sentence.
Show, demonstrate, mostrar and demostrar in medical articles 41

2.2 Definition of the environments

For contrastive analysis, whether theoretical or practical, it is necessary to set up


a framework for comparison, or tertium comparationis, that is not biased in one
direction or the other. This framework is usually established at a higher level of
abstraction than the phenomena under direct observation. For the four verbs con-
sidered here, the conceptual framework selected is that of the semantico-syntactic
environments in which they perform their rhetorical functions an approach
that has been successfully employed with the verbs report (Williams forthcoming)
and observar (observe) (Williams 2007). The categories for these environments
were based on relevant meanings of the verbs identified in the Collins Cobuild
English Language Dictionary (1987) and the Diccionario de la Real Academia Es-
paola (Spanish Royal Academy Dictionary) (2001). The relevant definitions are
shown in Table2.
Surprisingly, definition 1 for show did not include a human agent (cf. defini-
tion 1 for demonstrate). For some rare cases, it is also necessary to apply definition

Table2. Relevant dictionary definitions of the verbs under study as used in the medical
research article
English Spanish
Show Mostrar
1. If something shows that a state of affairs 1. Explicar, dar a conocer algo o convencer de su
exists, it gives information or evidence that certidumbre (Explain, make something known
proves it or makes it clear to people. or convince others of its truth or certainty.)
2. If something shows a quality, characteris- 2. Hacer patente un afecto real o simulado.
tic, or feature, there is evidence of it which Dar a entender o conocer con las acciones un
can be observed. If you show a particular calidad del nimo (Make clear a real or feigned
characteristic or feature, you behave in a feeling. Imply or make a quality clear by ones
way that makes it clear to people. actions.)
3. If you show someone affection, mercy, 3. (Dicho de una persona) Portarse de cierta
respect, etc, you behave towards them in manera, o darse a conocer en algn sentido
that way. ((Said of a person) Behave in a certain man-
ner, or make this clear in some way.)
4. If a picture, diagram, table, piece of writ- 4. Manifestar o poner algo a la vista (Display
ing, etc shows something, it represents it or or put something on view.)
expresses information.
Demonstrate Demostrar
1. If someone or something demonstrates a 1. Probar, sirvindose de cualquier gnero de
fact, theory, or principle, they make it clear demostracin (Prove by means of any type of
to people. evidence.)
2. If you demonstrate a particular skill, qual- 2. Manifestar, declarar (Manifest, declare.)
ity, or feeling, you reveal that you have it.
42 Ian A. Williams

2 for demonstrate to the behaviour and functioning of certain living cells and or-
gans (cf. definition 2 for show). In general, the Spanish definitions seem to imply
an active animate agent, but there is no reason why these cannot be applied to
inanimates or to be impersonalised through passivisation. In particular, definition
4 for mostrar implies: if someone displays something in a table, it is shown in the
table and the table shows it.
All the instances of the verbs were located in the three subcorpora with the
concordance tool of WordSmith Tools (Scott 1998). Occurrences of these verbs
were classified according to the dictionary definitions (Table2) into the following
categories: Authors, Techniques, Evidence, Metatextual and Characteristics. In
line with definition 1 for each of the four verbs and definition 2 for demostrar, the
category Authors includes all contexts, active and passive, in which an explicit hu-
man agent provides information or evidence that proves a state of affairs exists:1
(1) a. In 1960 Barritt and Jordan1 showed that anticoagulation reduced
mortality in pulmonary thromboembolism.
b. Lehman16 demostr que la desecacin produca exudado y favoreca la
coagulacin.
(Lehman16 demonstrated that desiccation produced exudate and
favoured coagulation.)

Techniques covers all active contexts in which application of a procedure, such as


a diagnostic imaging technique or statistical test, reveals or discloses that a state
of affairs exists:
(2) a. In the patients in this report contrast studies usually showed a caecal
abnormality.
b. El anlisis de la mortalidad [] no ha mostrado diferencias significativas
en los dos perodos analizados (6,4 frente al 7,71%).
(The mortality analysis [] showed no significant difference between
the two periods analysed (6.4 vs. 7.71%).)

Here, according to definition 1 for each of the verbs, the technique is the instru-
mental subject (pseudoagent) of the verb, with the human agency involved in the
process remaining hidden.
Through the same definitions, the Evidence category includes all active con-
texts in which the source of the evidence is expressed in the guise of research or
the results of such investigation, and again any human agency remains implicit:
(3) a. Recent studies have demonstrated the photodestruction of a
photosensitizer during treatment (photobleaching).68
b. Los resultados de este estudio muestran que las infecciones por Hib son
causa habitual de patologa grave en la edad peditrica.
Show, demonstrate, mostrar and demostrar in medical articles 43

(The results of this study show that Hib infections are a common cause
of serious illness in the pediatric age.)

The Metatextual category, according to definition 4 for both show and mostrar,
covers all contexts, both active and passive, with reference to figures and tables:
(4) a. Table II shows associations of hip osteoarthritis with height and weight.
b. Las nicas variables con una relacin estadsticamente significativa se
muestran en la tabla 6.
(The only variables with a statistically significant relationship are shown
in table 6.)

The category Characteristics refers to instances in which an entity, human or in-


animate, displays a characteristic, this being made apparent or evident without
necessarily mentioning any external influence as occurs in the preceding catego-
ries:
(5) a. After the second operation, none of the patients showed significant
intraocular pressure elevations.
b. The other biochemical parameters showed the same pattern.
c. En general el conducto de Schlemm muestra una luz abierta y amplia.
(In general, Schlemms canal shows a broad open lumen.)

This category is based on definitions 2 and 3 for show, definition 2 for demonstrate,
definitions 2 through 4 for mostrar, and definition 2 for demostrar.
In addition to the semantic categories, categories based on syntactic struc-
ture were also considered because, with the exception of the Characteristics cat-
egory, the defining semantic components may be deliberately suppressed, can be
expressed by other means, particularly as adjuncts, or cannot be identified. The
presence of any of these factors results in an impersonal statement and, although
there are similarities between English and Spanish, systemic differences entail
separate presentation in each case. For English, the two syntactic categories were
Passives and Hybrids. The category Passives included what can be considered
pure passives, either with the phenomenon as subject of a passive verb plus any
other complementary material, as in (6a) and (6b), or with dummy subject it and
a postposed projected clause (6c):
(6) a. The benefits of endoscopic surgery were clearly shown in this study.
b. A definitive diagnosis of TB can be made only if caseation is
demonstrated on histological examination, or if acid-fast bacilli are
demonstrated in the tissue by Ziehl-Nielsen staining, culture, or guinea-
pig inoculation.
44 Ian A. Williams

c. Furthermore, it has been shown that in a nutritionally at-risk


population, the level of motor maturation at 15 months predicts
cognitive test performance at 18 years of age.18

The hybrid forms, which are active or stative sentences with an added passive as-
pect, take the form of passive show plus infinitive. They are often equivalent to
the impersonal passive structures with dummy it illustrated in (6c), since they
also allow suppression of the agent, whether human or expressed metaphorically
by a technique or evidence; however, they constitute a resource for establishing
new themes in the discourse or for developing thematic progression through topic
continuity:
(7) Preoperative DI [detrusor instability] was shown to be a clear risk factor
for the development of postoperative DI, increasing the risk significantly
(Table II).

In (7), the meaning is We/The statistical analysis/The study showed that preop-
erative DI is/was a clear risk factor, but the structure allows Preoperative DI to
be foregrounded as theme.
For the Spanish texts and the translations, the Passive category included in-
stances of the true passive ser + past participle (8a), pseudopassives with estar or
quedar + participle (8b), and the so-called reflexive passive with se (8c), which is
a middle construction that is focused on the event with all connotation of agency
suppressed (Nava & Maldonado 2004:462):
(8) a. El valor del ketoconazol en las dermatofitosis ha sido demostrado en el
manejo de estas enfermedades en algunos estudios abiertos.46
(The value of ketoconazol in dermatophytoses has been demonstrated
in the management of these diseases in some open trials.46)
b. No est concluyentemente demostrado que la supervivencia del cncer
de esfago mejore con la realizacin de amplias linfadenectomas
mediastnicas transtorcicas.2,34,35
(It has not been [lit. is not] conclusively demonstrated that survival in
oesophageal cancer is improved by performing extensive transthoracic
mediastinal lymph node resection.2,34,35)
c. La distribucin normal de los datos se demostr por la prueba de
Kolmogorov-Smirnov.
(Normal distribution of the data was demonstrated by the
Kolmogorov-Smirnov test.)

A further category termed Reflexive was used to cover structures similar in func-
tion to the English Hybrids but not formally passive. For mostrar the structure
Show, demonstrate, mostrar and demostrar in medical articles 45

consists of reflexive mostrarse plus an adjectival complement (9a) or a preposi-


tional complement with como (9b):
(9) a. En nuestra serie la ETE se mostr muy superior a la ETT en la deteccin
de las regurgitaciones tanto en las prtesis normofuncionantes como en las
disfuncionantes.
(In our series transesophageal echocardiography was shown to be
[lit. showed itself] far superior to transthoracic echocardiography in
detecting regurgitation both in normal and dysfunctional prostheses. )
b. La DMTU es un saturador especfico de radicales hidrxilo y se muestra
como un agente antiinflamatorio muy efectivo.6
(Dimethyl-thiourea is a specific hydroxyl radical scavenger and
has been shown to be [lit. shows itself as] a highly effective anti-
inflammatory agent.6)

These are also middle structures in which the event is portrayed as taking place
within one participant. The reflexive se does not imply the dual role of Agent-Af-
fected, but suppression of the notion of agency (Nava & Maldonado 2004). In nei-
ther (9a) nor (9b) could the se form be replaced by the agentive passive with ser.
In the case of demostrar, the syntactic structure with the equivalent reflexive
meaning is not expressed formally through the reflexive se form but through syn-
tactic extension of the verb phrase, expressed by active demostrar plus an infini-
tive, and auxiliarisation of the finite verb:
(10) Slo dos frmacos virostticos han demostrado ser efectivos en la retinitis por
CMV administrados por va sistmica, el ganciclovir y el foscarnet.
(Only two virostatic drugs have proved [lit. demonstrated] to be effective in
cytomegalovirus retinitis when administered systemically: ganciclovir and
foscarnet.)

Semantically, there is a shift of responsibility for the action from an implied ex-
ternal human agent to the inanimate entity. This is seen to manifest a particu-
lar quality through its chemical behaviour, thereby making the event internal to
the subject, similar to the more dynamic representations in the Characteristics
category.
Both mostrar and demostrar very occasionally take a projected clause in which
the subject of the dependent verb is the same as that in the projecting clause (i.e.,
reflexive in meaning):
(11) La utilizacin de dexametasona en las meningitis por Haemophilus influenzae
tipo b ha mostrado que disminuye los niveles de lactato y protenas del lquido
cefalorraqudeo, eleva la glucorraquia y previene la aparicin de sorderas.2729
46 Ian A. Williams

(Use of dexamethasone in Haemophilus influenzae type B meningitis has


been shown to reduce [lit. has shown that it reduces] cerebrospinal fluid
(CSF) lactate and protein levels, to increase CSF glucose, and to prevent
deafness.2729)

2.3 Quantitative and contextual analyses

The subcorpora were compared with regard to the frequency of the verbs over-
all and within the different categories. Within these environments, the contextual
analysis examined different discoursal, syntactic or semantic features which have
been found relevant for lexical and syntactic choice in previous studies (Williams
1996b, 1999, 2005, 2006, 2007). These include rhetorical function (RPR, self-ci-
tation, SOR, claim) for all categories, first person versus third person subject for
Authors, clausal vs. noun direct object for Authors, Techniques and Evidence,
active vs. passive for the Metatextual category, subject type for Techniques and
Characteristics, object type for Characteristics, and type of passive in the Pas-
sive category. Statistical analysis was performed with SPSS (version 14). Verb fre-
quencies were compared with a binomial distribution test. Raw frequencies were
normalised per 100,000 words for representation in the figures. For the categori-
cal variables, 2 n contingency tables were constructed and the subcorpora were
compared with the test. A P value of 0.05 was considered significant.

3. Results

3.1 Overall frequencies and distribution of the verbs

In ESLC there were 299 tokens of show and 69 tokens of demonstrate, a ratio of 4:1.
In SNLC, mostrar appeared 91 times compared with 116 instances of demostrar.
The figures for these verbs in STLC were 181 and 163 tokens, respectively. If the
two verbs in each subcorpus are taken together, the two English verbs are found
to occur more frequently, with 234 tokens per 100,000 words compared with 147
instances in SNLC and 186 in STLC.
The normalised data for the individual verbs expressed per 100,000 words are
given in Table3. Since there is no one-to-one relation between the English and the
Spanish verbs, the translation process is seen to lead to an overall excess of mostrar
in the TL texts (p<0.001 by the binomial test), whereas total use of demostrar is
almost identical in STLC and SNLC.
Table4 gives the raw frequencies for the verbs in the different semantic and
syntactic categories.2 For show, each of the three main categories Characteristics,
Show, demonstrate, mostrar and demostrar in medical articles 47

Table3. Overall normalised frequency of the verbs under study in the three subcorpora
ESLC SNLC * STLC * ESLC SNLC STLC
show mostrar mostrar demonstrate demostrar demostrar
Frequency
per 100,000 190 65 98 44 83 88
words
Note: * p<0.001 by the binomial distribution test; not statistically significant

Table4. Frequencies of the verbs under study in the three subcorpora by semantic and
syntactic category
ESLC SNLC STLC
Category Show Demon Total Most Demos Total Most Demos Total
Authors 23 13 23 0 21 15 4 25 16
Techniques 37 6 27 15 19 24 27 13 22
Evidence 62 24 55 10 27 26 28 55 45
Metatextual 66 1 42 9 1 7 44 1 24
Characteristics 73 5 49 48 2 36 63 3 36
Passive 11 19 19 1 33 24 2 48 27
Hybrid/Reflexive 19 0 12 8 5 9 7 14 11
Others 8 1 6 0 8 6 6 4 5
Total 299 69 233 91 116 147 181 163 186
Note: Data for the individual verbs are raw frequencies, whereas the combined totals are normalised and
expressed per 100,000 words. Demon=demonstrate; Most=mostrar; and Demos=demostrar

Metatextual and Evidence has more or less the same frequency as demonstrate
overall, and these are followed at some distance by Techniques and Authors. Show
predominates over demonstrate in all the semantic environments, and almost ab-
solutely in the Metatextual and Characteristics categories. Of the syntactic cat-
egories, the Hybrid forms of show outnumber the Passives by about 2 to 1. For
demonstrate, the category with the highest frequency is Evidence, accounting for
1 of every 3 instances, followed by Authors, the other semantic categories being
uncommon with this verb. Unlike show, demonstrate is often associated with the
pure passive, this being the only function in which demonstrate exceeds show.
In SNLC, the outstanding function of mostrar is to express Characteristics,
with over half the tokens falling into this category. This is followed by Techniques,
Evidence and Metatextual. No instances of mostrar with researcher as subject
were attested in the Spanish texts. Of the syntactic categories, the reflexive use is
not common, but the passive is rare; the single instance was a middle structure
(se form) representing an inagentive almost intransitive use rather than a genu-
ine passive. Spanish demostrar shows a fairly close parallel with its English coun-
terpart, the most frequent category being Evidence, followed by Authors and
48 Ian A. Williams

35

30
P<0.001
25
Per 100,000 words

20
P=0.014
15

10

es
rs

s
es

l
nc

ve
s
ua
o

iv
tic
qu
th

xi
e

ss
t
ex
id

ris
Au

ni

fle
Pa
Ev

at
ch

te

Re
et

ac
Te

ar
Ch

Spanish Translation
Figure1. Comparison of use of the verb mostrar in the Spanish and Translation subcor-
pora by semantic and syntactic category.

Techniques, while use related to figures and tables or to express characteristics is


rare in the texts analysed. Like demonstrate, demostrar frequently appears in the
passive, but also displays the additional reflexive function as an occasional struc-
tural and thematic variant of the se-type passive with a postposed clausal subject.
In STLC, the most frequent use for mostrar again corresponds to Character-
istics but this is followed by a surprisingly high figure for the Metatextual cat-
egory, with Evidence and Techniques having similar frequencies. Four instances
of mostrar appear with researchers as subject or agent. As in SNLC, the Reflexive
syntactic category outnumbered the Passive, which was rare with mostrar. Fig-
ure1 compares the normalised data for mostrar in SNLC and STLC. Use of this
verb for Authors may be unnatural, but the outstanding results are the significant
excesses in the Evidence (p=0.014) and Metatextual (p<0.001) categories.
For demostrar, the pattern of use is not unlike that seen in SNLC, the outstand-
ing semantic category being Evidence, followed by Authors and Techniques, with
only rare use of the verb to express characteristics or in metatextual reference to fig-
ures and tables. The pattern for passive and reflexive forms also parallels that in the
Spanish texts. The normalised data for SNLC and STLC are compared in Figure2,
which shows a small but non-significant excess for Reflexive, a significant deficit
for Techniques (p=0.026), and a significant excess for Evidence (p=0.015).
Show, demonstrate, mostrar and demostrar in medical articles 49

35
P=0.015
30

25
Per 100,000 words

20
P=0.026
15

10

s
rs

s
es

ive

ve
c

ics
a
o

en

tu
qu
th

xi
ss
t
ex

ris
id
Au

ni

fle
Pa
Ev

at

te
ch

Re
et

ac
Te

ar
Ch

Spanish Translation
Figure2. Comparison of use of the verb demostrar in the Spanish and Translation subcor-
pora by semantic and syntactic category.

3.2 Cross-linguistic comparison by category

The data in Table4 can also be interpreted horizontally for each of the semantic
and syntactic categories. In the category Authors, joint use of the two English
verbs (23 instances per 100,000 words) exceeds that of Spanish (15/100,000), and
show outnumbers demonstrate by about two to one. In SNLC and STLC, this func-
tion corresponds almost exclusively to demostrar, and the overall frequencies are
similar (15 and 16 per 100,000, respectively).
For Techniques, combined use is slightly higher in ESLC (27/100,000 words)
than in SNLC (24/100,000). However, whereas the English texts display a prefer-
ence for show, the figures are fairly balanced in the Spanish texts. In STLC, com-
bined use is a little lower (22/100,000) than in SNLC despite the slightly higher fig-
ures in ESLC. For the individual verbs, the TL texts reflect a preference for mostrar
over demostrar, although this is not so pronounced as that of English show.
The Evidence category shows the highest combined use of show and demon-
strate (62 and 24 tokens, or 55/100,000 words), whereas this function takes second
place in SNLC, at 26/100,000 (mostrar 10 tokens and demostrar 27). In STLC, the
corresponding figure is 45/100,000 (mostrar 28 tokens and demostrar 55). Although
the proportional use of the two verbs is correct, approximately one instance of
mostrar for every two of demostrar, both verbs are clearly overrepresented for this
50 Ian A. Williams

function: if the SNLC figures of 10 and 27 tokens are taken as the statistical norm,
they predict 13 and 36 instances, respectively, for the slightly larger STLC.
Metatextual reference to tables and figures is the exclusive domain of both
show and mostrar in all three subcorpora, but the TL texts exhibit a pattern of use
that comes much closer to the SL text pattern than to that of the Spanish texts.
The verbs show and mostrar also predominate in the category Characteristics,
but in this case, despite the higher frequency in ESLC, the overall use and relative
frequencies in SNLC and STLC are almost identical (see Figures1 and 2).
For the syntactic categories, variation in use is predicted by the systemic dif-
ferences indicated in Section2.2. Thus, whereas both show and demonstrate are
used in the true passive, passive forms in Spanish, whatever their type, are ex-
pressed predominantly through demostrar, with a few instances of mostrar in the
se-type passive. Despite the systemic differences, the TL text pattern follows the
Spanish NL pattern closely (Figures1 and 2). With regard to English Hybrids
with show, only those with the infinitive to be + complement correspond with
Reflexive mostrar whereas all infinitive types (to be, to have, to reduce) can be ren-
dered by the active form of demostrar + infinitive (demostrar + ser/tener/reducir).
Although tokens of Reflexive mostrar outnumber those of demostrar in SNLC, the
reverse is true in STLC, where joint use is also slightly higher.

3.3 Analysis of associated contextual features

No statistically significant differences were found between the three subcorpora


for any of the semantico-syntactic categories in relation to rhetorical function.
Nevertheless, for the Evidence category, the similar distribution in the Spanish
NL texts and the TL texts indicates that the statistically significant excess observed
for mostrar and demostrar in STLC for this category is expressed in all the rhe-
torical functions. Similarly, there were no differences in the distribution of first-
person use and third-person reference in the category Authors, or for clausal vs.
noun DO in the Authors, Techniques and Evidence categories.
For Techniques, analysis of the nouns representing the technique or proce-
dure as pseudoagent subject showed no difference between the subcorpora. The
nouns could be grouped under the broad heads of diagnostic imaging techniques
(12a), other diagnostic tests or investigative procedures (12b), statistical tests and
analysis (12c), and cognitive activities and calculation (12d).
(12) a. Control intraoperative angiography showed patency of the small
intrahepatic arterial branches.
> La angiografa intraoperatoria de control mostr una permeabilidad de
las ramas arteriales intrahepticas pequeas.
Show, demonstrate, mostrar and demostrar in medical articles 51

b. Los estudios biomicroscpicos demostraron Tyndall hemtico en un caso,


no encontrndose en este grupo catarata asociada en ningn caso.
(Biomicroscopic studies demonstrated haematic Tyndall phenomenon
in one case, and in no case in this group was there an associated
cataract.)
c. La comparacin de la incidencia de hemorragias entre el primero y
segundo perodo ha mostrado una diferencia significativa (p =0,041) con
menor incidencia en los pacientes de la segunda poca (16,46 frente al
11,17%) (fig. 2).
(Comparison of the haemorrhage rates in the first and second periods
showed a significant difference (p=0.041), with a lower incidence in
patients in the second period (11.46 versus 11.17%) (fig. 2).)
d. Examination of this table shows that in patients with NL, the aggregate
length of cul-de-sacs is about half the length of the wall of SC.
> El examen de esta tabla muestra que en los pacientes NL, la longitud
acumulada de los fondos de saco es de alrededor de la mitad de la longitud
de la pared del CS.

For the Metatextual category, the contextual study was extended to cover all
the verbs expressing the pointer function in the three subcorpora. Table5 indi-
cates the lexical range for each subcorpus and shows the verbs appearing twice or
more, the category Others representing verbs used only once or translation by
non-verbal means. The English SL texts contained 23 different verbs covering a

Table5. Frequency of the lexical verbs used for metatextual reference to tables and figures
ESLC SNLC STLC
Show 66 Exponer 10 Mostrar 44
Present 10 Mostrar 9 Exponer 14
Give 9 Reflejar 9 Presentar 11
List 9 Recoger 8 Describir 9
Compare 8 Resumir 7 Resumir 9
Summarise 8 Representar 6 Comparar 8
Include 4 Detallar 5 Indicar 7
Illustrate 3 Observar 5 Incluir 5
Indicate 3 Apreciar 4 Ilustrar 4
Visualise 2 Describir 4 Representar 3
Others 13 Comprobar 3 Dar 2
Indicar 3 Detallar 2
Presentar 3 Enumerar 2
Aparecer 2 Observar 2
Expresar 2 Visualizar 2
Others 9 Others 13
Total 135 Total 89 Total 137
52 Ian A. Williams

total of 135 instances, the predominant verb being show with 66 tokens (49%). The
Spanish texts had 24 verbs for only 89 instances, and the maximum exponent was
exponer with 10 tokens, followed by mostrar with 9 (only 10% of the total). The TL
texts had 25 different verbs for 137 instances, and mostrar predominated with 44
tokens (32%). Use in the TL texts, therefore, comes closer to the English than to
the Spanish pattern.
Comparison of SNLC and STLC with regard to the verbs used at least twice
in either of these subcorpora showed that 11 verbs were common to both lists. Of
these, exponer (show), resumir (summarize) and indicar (indicate) appeared at
equivalent frequency; mostrar (show), comparar (compare), presentar (present)
and describir (describe) showed excesses in the TL texts, whereas recoger (gath-
er), representar (represent), detallar (detail) and observar (observe) were all
underrepresented. In addition, five verbs (incluir (include), ilustrar (illustrate),
dar (give), enumerar (list), visualizar (visualise)) appeared in STLC but not in
SNLC. Similarly, five verbs (reflejar (reflect), apreciar (see), comprobar (verify),
aparecer (appear), expresar (express)) occurred in SNLC but not in STLC. The
most significant deficits were those corresponding to reflejar, which occurs with
both tables and figures, and recoger, plus the set of observation verbs observar,
apreciar and comprobar, which appear in subordinate clauses attached to a main
clause expressing a result or a comment on the content of the figure or table:
(13) La HTA no se asocia significativamente con la hipercolesterolemia (C >200
mg/dl) pero s con la hiperlipidemia cuando est definida por LDL/HDL >2,2,
como puede comprobarse en la tabla 1.
(Hypertension is not significantly associated with hypercholesterolemia (C
> 200 mg/ dL) but it is associated with hyperlipidemia when this is defined
as LDL/HDL > 2.2, as can be seen in table 1.)

The gloss is literal, but a briefer more natural version would be as shown in ta-
ble1.
A second notable feature identified in the contextual study is the syntactic
structures used in this rhetorical function. In the English texts, subject + active
verb occurs in 44 instances (33%), which contrasts with the total of 4, a mere 5%,
in the Spanish texts (p<0.001 by analysis). In the TL texts, this syntagm was
found in 37 contexts (27%, p=0.006, when compared with the Spanish texts), thus
reflecting the SL text pattern rather than the TL usage.
Whether the noun (table or figure) was expressed as the grammatical subject
or in an adjunct, these components were thematic in 57 of 135 contexts (42%) in
ESLC whereas thematicity was slightly higher in SNLC, at 41 of 89 instances (46%,
not statistically significant). In STLC, however, the corresponding figure was 91 of
137 instances (66%, p=0.004, when compared with SNLC). A total of 33 rhematic
Show, demonstrate, mostrar and demostrar in medical articles 53

metatextual references in the SL texts became thematic in the TL texts. The single
instance of the inverse movement occurred in conjunction with a rheme-theme
exchange:
(14) The same risk estimates as those shown in the previous table are repeated
in Table V, except that here the definition of SGA birth is changed to include
only those newborns with a birth weight below the nominal 10th percentile.
> En la Tabla V se repiten las mismas estimaciones sobre el riesgo que en la
tabla anterior, excepto que en sta, la definicin de hijos PEG se cambia para
incluir slo los recin nacidos con un peso al nacer por debajo del percentil 10.
(Table V repeats the same risk estimates as the previous table, except that
[])

The reason for the consistency of this thematization in the translations is not clear
but may be related to a preference to place the heavier component in clause-final
position based on the principle of end weight, and this may be done regardless of
the dynamics of the progression of the discourse, and the counter principle of end
focus.
For the category Characteristics, analysis of the entities characterised (i.e. the
subject headnoun) showed that the English texts and the translations tended to
have more nouns representing humans, either individually (e.g. patients), collec-
tively (groups) or in the abstract (cases) than did the Spanish texts: 29 of 78 in-
stances (37%), 24 of 66 (36%), and 10 of 50 (20%), respectively, but the difference
did not quite reach the level of statistical significance.
In a qualitative analysis, the collocates of show in subject position were patient/s
(13 tokens), children (2), infant/s (2), survivors (2), and women (1 token); collec-
tive nouns were represented by group/s (3 tokens), subgroup (1), and subset (1),
and there was one instance of case representing a human subject The DZP exposed
case. In contrast, of the 10 instances of human subjects in the Spanish texts, 4 were
represented by the abstract casos, there were 3 tokens of madres (mothers), 2 of
escolares (schoolchildren), with one instance of a collective poblacin con bocio
(population with goitre). While the differences are apparent, it is difficult to draw
conclusions from such a small sample.
For the non-human subjects, the lexical items collocating with show and
mostrar as grammatical subject are compared in Table6. Overall, ESLC and SNLC
display considerable similarity both quantitatively and in the wide range of enti-
ties covered, with most concepts classifiable under seven broad headings: types of
sample, microscopic entities, parts of body, clinical entities, methods and proce-
dures, measurable phenomena, and physiological entities. Only the English texts
expressed the characteristics of data or evidence: e.g. the variability of cardiac event
rates. On the other hand, only the Spanish texts characterised substances, both
54 Ian A. Williams

Table6. Non-human lexical items collocating with show and mostrar as grammatical
subject
ESLC SNLC
Types of sample 8 Types of sample 5
Biopsies 2 Biopsias (biopsies) 1
Sections 1 Espcimen (specimen) 1
Slides 2 Grupos (de clulas) (groups (of cells)) 3
Specimens 3
Microscopic entities 7 Microscopic entities 8
Cells 3 Clulas (cells) 3
Distribution (infiltrate) 1 Conducto de Schlemm (Schlemms canal) 2
Epithelium 1 Estructura del ngulo (structure of the angle) 1
Nests of cells 2 Leucocitos (leucocytes) 1
Trabcula (trabecula) 1
Parts of body 10 Parts of body 6
Epidermis 1 Epidermis (epidermis) 1
Eyes 4 Nevos (nevi) 1
Leaflets 1 Piel (skin) 2
Areas (of leaflets) 1 Prtesis (prosthesis) 1
Portion (of leaflet) 1 Superficie (piel) (surface (skin)) 1
Nevus/ nevi 2
Clinical entities 5 Clinical entities 3
Conditions 1 Adhesiones (adhesions) 1
Cancer 1 Tumores (tumours) 2
Carcinoma 1
Histiocytosis X 1
Lesions 1
Methods and procedures 5 Methods and procedures 3
Determinations 1 Determinaciones (determinations) 2
Radio /chemotherapy 1 Histologa (histology) 1
Radiological methods 1
Laser treatment 1
Loco-regional treatment 1
Measurable phenomena 7 Measurable phenomena 5
Changes 3 Caractersticas (characteristics) 1
Differences 1 Distribucin estacional (seasonal distribution) 2
Parameters 1 Sistema de valoracin (assessment system) 1
Relative risks 1 Perodo postoperatorio (postoperative period) 1
Values 1
Physiological entity 1 Physio/psychological entities 2
Ejection fraction 1 Cumplimiento (compliance) 1
Respuesta micolgica (mycological response) 1
Data or evidence 4 Natural/artificial substances 5
Cine frames 1 Frmacos (drugs) 1
Data 1 Sustancias antioxidantes (antioxidative agents) 1
Study/ studies 2 Complejo E-1-PI (E-1-PI complex) 1
Elastasa (elastase) 1
Lquido cefalorraqudea (cerebrospinal fluid) 1
Animals 1
Grupos (de cerdos) (groups (of pigs)) 1
Show, demonstrate, mostrar and demostrar in medical articles 55

Table7. Lexical items collocating with show and mostrar as direct objects
ESLC SNLC
Visible phenomena (n = 33) Visible phenomena (n = 19)
abnormal caecum anomala (estructural) (structural anomaly)
alteration (to elastic fibers) cambios histolgicos (histological changes) (3)
atypia color (de piel) (colour (of skin))
autoantibodies configuracin (configuration)
bulla (intraepidermal) distrofa miotnica (myotonic dystrophy)
cancers (second primary) fibras elsticas (elastic fibres)
cells lesin ulcerosa (ulcerating lesion)
changes (histological) luz amplia (broad lumen) (2)
deposits (3) morfologa (morphology)
detrusor instability positividad (positivity)
diameter (on cine frame) queratinocitos (keratinocytes)
dysplasia shock y CID (shock and DIC)
eosinophilic spongiosis signos de alteracin / afectacin
enlargement (left ventricle) (signs of alteration / involvement)
episode of desaturation (on recording) ultraestructura (ultrastructure)
overlapping features vascularizacin (vascularisation)
fibrosis
granules
hyperplasia (2)
lesions
motion (3)
nucleus
apneic pauses (on recording)
pemphigus foliaceus (on slide)
picture (clinical)
scarring
signs of life
similarities (cytological)

natural and artificial, with mostrar, and one text described groups of experimental
animals with this verb. These differences in lexical choice most probably reflect
differences in study design and the type of data generated in them.
Table7 lists the entities expressing the characteristic in the direct object posi-
tion. These were found to fall into two broad groups, visual phenomena and ab-
stract entities. There was no statistically significant difference in the distribution
of the two types between the subcorpora, with abstractions slightly exceeding the
visible phenomena in all three subcorpora.
A more detailed analysis of the entities in the two groups revealed certain se-
mantic affinities. Thus, most abstractions were those obtained by laboratory anal-
ysis (concentraciones, niveles, valores) or by statistical methods (differences, ten-
56 Ian A. Williams

ESLC SNLC
Abstractions (n = 40) Abstractions (n = 29)
activity (receptor) accin antinflamatoria (antiinflammatory action)
age (average) aumentos (increases)
changes (3) capacidad (capacity) (2)
criteria (clinical) concentraciones (concentrations) (2)
decrease diferencias (differences) (5)
deterioration (4) disminucin (decrease)
differences (2) excrecin de yoduria (iodine excretion)
discrepancies fase de sntesis (synthesis phase)
effects (positive) incremento (increase)
elevations (2) inters (interest)
events (number) media de talla y peso (mean height and weight)
improvement (3) niveles (levels)
increase (3) porcentajes de bocio (percentages of goitre)
index predictividad (predictivity)
maturation predominio (predominance) (2)
pattern recuento (count)
promise rentabilidad (cost-effectiveness)
outcomes (positive/negative) sensibilidad (sensitivity)
rate tendencia (tendency) (2)
reduction valores (values) (2)
risks (2)
scores
sensitivity
tendency
variability (2)
variation

dency, risks; diferencias, tendencia), and those reflecting changes in level (decrease,
elevations, increase, reduction; aumento, disminucin, incremento) or changes in
state (deterioration, improvement). Among the visible phenomena, there is a pref-
erence for those observed at microscopic level (histologic changes, cambios his-
tolgicos), or revealed by imaging or staining techniques (abnormal caecum, left
ventricular enlargement, positividad para HKN-5) or made visible by special equip-
ment (episodes of desaturation, apneic pauses, these being recorded on monitoring
equipment).
It is noteworthy that few instances of show or mostrar collocate with phe-
nomena evident at the macroscopic level or directly observable by the clinician.
Combination is possible when the subject is inanimate: both conditions may show
flexural papular lesions, and mucocutaneous lesions, and in one context in which
the physical uncovering of a lesion was described:
Show, demonstrate, mostrar and demostrar in medical articles 57

(15) Los cuatro perforantes tenan una escamocostra, que al desprenderse muestran
una lesin ulcerosa.
(The four perforating nevi had a scaly scabby covering and when this
sloughed off showed an ulcerating lesion.)

However, outwardly visible phenomena do not combine readily with human sub-
jects, which would create ambiguity between the function of mostrar to describe
characteristics with a weak Possessor as agent, and its function as an action verb
with a dynamic controlling Agent, possibly requiring the presence of an indirect
object Receiver. In SNLC, one instance of a disease with overt features occurred
in combination with an abstraction:
(16) Ninguno de estos casos mostraba distrofia miotnica ni historia familiar de la
misma.
(None of these cases showed myotonic dystrophy or a family history of the
disease.)

A second example, shock y CID (coagulacin intravascular diseminada (dissemi-


nated intravascular coagulopathy)) combined a condition with visible manifesta-
tions with one requiring specific laboratory analyses. Combinations of this type
remove the ambiguity by ruling out the interpretation of the subject as a control-
ling Agent. Both examples occurred with casos, and the role of the more abstract
cases as opposed to patients in such contexts is uncertain based on just two in-
stances.
Another two contexts concerning macroscopically observable phenomena are
of interest, since they may involve items that can always collocate with show and
mostrar. The first contains the mention of colour:
(17) La superficie que lo recubre habitualmente muestra el color de la piel
circundante.
(The surface covering it usually shows the colour of the surrounding skin.)

The second context includes the lexical item signos (signs), of which there were
two instances in SNLC and one in ESLC. For both colour and signs, the interpreta-
tion is based on intuition, and needs to be verified in a larger sample.
The syntactic structures involving Passives and Hybrids in English and Pas-
sives and Reflexives in Spanish are essentially impersonal statements that allow
the agent to be suppressed.3 They are also variants of the active forms in which
human agency is disguised behind the techniques or studies, and serve to achieve
different effects in thematization and information flow. However, techniques and
the source of the evidence (studies, data, etc) are sometimes included as adverbs
(e.g. histologically), as adjuncts expressing means (e.g. by Ziehl-Nielsen staining)
58 Ian A. Williams

or location (e.g. in epidemiological studies of patients with Hodgkins disease). The


contextual study using the categories Means, Data source, or None showed that
such references were present in approximately half of the statements, but no differ-
ences were found between the subcorpora in their distribution.
One striking feature of the TL text passives is the predominance of the middle
se type (42 instances vs. 21 in the Spanish texts) over other possible variants: true
passive with ser (5 instances vs. 7), and descriptive passives with the substitute
verbs estar (1 instance vs. 3) and quedar (0 vs. 2 instances). Numbers were too
small to carry out an analysis including all the categories separately, but compari-
son of the se form against all other types for demostrar showed a significant differ-
ence between STLC and SNLC (=5.136; 1 df; p=0.023).
The variant forms with ser, as in example (18), and quedar, in example (19),
are indicated to express claims and RPRs in association with the present perfect
tense and with an adjunct acting as a kind of pseudoagent:
(18) Los pacientes tratados con dipiridamol tericamente podran tener una
respuesta excesiva a la administracin de ATP, ya que se ha descrito en
preparaciones experimentales que el frmaco interfiere en su catabolismo8.
Clnicamente este hecho no ha sido demostrado.
(Patients treated with dipyridamole could theoretically show an excessive
response to ATP administration, since in experimental preparations the
drug has been reported to interfere with its catabolism.8 Clinically this has
not been demonstrated.)
(19) Segn nuestros resultados ha quedado demostrado un aumento del
infiltrado corneal en algunos grupos problema tratados con saturadores
de radicales libres como SOD y DMTU, es decir empeoran la queratitis
infecciosa.
(According to our results has been demonstrated an increase in the corneal
infiltrate in some problem groups treated with free radical scavengers like
SOD and DMTU, that is, they make the infectious keratitis worse. [i.e. Our
results have demonstrated an increase ])

The variant with estar is used to express consensus statements either in the Intro-
duction or the opening of the Discussion section:
(20) Est claramente demostrado que la angioplastia produce una mejora de los
sntomas, aumenta la capacidad funcional de los pacientes y eleva el umbral de
isquemia4,10,26,31,32.
(It is [has been] clearly demonstrated that angioplasty improves
symptoms, increases patients functional capacity, and raises the ischemic
threshold.4,10,26,31,32)
Show, demonstrate, mostrar and demostrar in medical articles 59

It is significant that all instances of passives with ser in both SNLC and STLC, and
those with quedar, were associated with an adjunct expressing either the tech-
nique or the source of the evidence. In contrast, few of the Hybrids and Reflex-
ives, which are clearly middle structures expressing a subject-internal process, had
these associated circumstancial elements. The se-type structures fall in between, so
that for SNLC we may speak of an agentivity gradient, with the ser and quedar pas-
sives conveying an implicit agent, with the se type focusing on the event, and with
the semantically Reflexive or internal forms as the least agentive and expressing
characteristics through their weak Possessor-type subject.

4. Translation behaviour

This section describes the way the verbs show and demonstrate have been man-
aged in the translation process both overall and within the different semantic and
syntactic categories. It reflects not only the relationships of the two SL verbs with
mostrar and demostrar but also with other verbs within their semantic scope that
translators feel the need to use. From the TL viewpoint, it will also reveal what
sources give rise to mostrar and demostrar other than the English verbs under
study.
Table8 shows the data on the transfer of show and demonstrate into the TL
texts by different means: namely, by mostrar in the same form, by demostrar in
the same or different category, by other verbs, or by non-verbal means. The right
hand column indicates the percentage cover by mostrar and demostrar (i.e. first
three columns) in relation to the row totals. Overall, we can see that of 368 tokens
of show and demonstrate, 135 were translated by mostrar, 115 by demostrar in the
same form and 18 by a different form of this verb, these 268 tokens accounting for
73% of the total.
The data reveal that translation of show involves a greater variety of choices,
since it is not only the major source of both mostrar (133 tokens) and demon-
strate (72 tokens) but is also transferred by other verbs (84 instances) or means
(10 times). Coverage of show by the two Spanish verbs is 69%. Within the cat-
egories, the pattern varies, with the greatest consistency and highest coverage
in the Evidence (84%) and Hybrid (84%) categories, and the greatest variety
and lowest coverage in the human Characteristics category (35%), which was
separated from the inanimate Characteristics category (70%) precisely because
of this different behaviour. In contrast, demonstrate can be translated consistently
by demostrar, with only two tokens being translated by mostrar, and another six
by other verbs.
60 Ian A. Williams

Table8. Translation replacements of show and demonstrate by semantic and syntactic


category.
Show
Mostrar Demos- Demos- Other Non- Row Cover*
trar trar Verbs Verbal Total %
Same Other
form form
Authors 4 13 6 23 74
Techniques 19 6 10 2 37 68
Evidence 25 24 3 10 62 84
Metatextual 39 1 25 1 66 61
Characteristics (human) 9 16 1 26 35
Characteristics (inanimate) 30 1 2 10 4 47 70
Passives 1 7 3 11 73
Hybrids/ Reflexive 1 7 8 1 2 19 84
Others 5 3 8 63
Subtotal 133 59 13 84 10 299 69
Demonstrate
Authors 12 1 13 100
Techniques 2 2 2 6 67
Evidence 23 1 24 100
Metatextual 1 1 100
Characteristics 1 1 1 1 3 67
Characteristics 2 2 2 100
Passives 16 3 19 84
Others 1 1 100
Subtotal 2 56 5 6 69 91
Total 135 115 18 90 10 368 73
Note: *Cover is expressed as a percentage of the sum of the first three columns in relation to the cor-
responding row total.

4.1 Translation in the Authors category

In the Authors category, all 13 tokens of demonstrate were translated by dem-


ostrar, 12 with identical function and the other changing from a first-person form
to the impersonal se-type passive. In addition, 13 tokens of show were translated
by demostrar as predicted by the SNLC data, which indicate that presentation of
strong evidence through a human agent is with this verb not mostrar. However,
there were four TL text instances of mostrar derived from show. One case appears
to be exceptional for this register:
(21) For comparison, we also show the appearance of two markedly affected
specimens from patients with PG (Fig 3, middle) and POAG (Fig 3, bottom).
Show, demonstrate, mostrar and demostrar in medical articles 61

> Para comparacin, mostramos tambin la apariencia de dos especmenes


marcadamente afectados de pacientes con GP (fig 4, centro) y GPAA (fig 4,
inferior).
This metatextual use with overt human agency was not attested in the Spanish
subcorpus, but is in line both with dictionary definition 4 (Table2) and with use in
medical texts of a didactic nature, which include a considerable amount of graphic
information. The other three cases can be regarded as deviant, and could be ren-
dered better either by demostrar or by other verbs as illustrated below in examples
(22) and (23).
Of the 6 instances of show translated by other verbs, 5 were expressed in a
more neutral tone: observar (observe) (twice), describir (describe), hallar (find)
and informar (inform), the exception being sealar (point out). These contexts,
together with those in which mostrar was used instead of the canonical demostrar,
contained certain circumstances that may condition this downtoning: the need
for modesty in self-citation or in presenting a current claim (3 instances), implicit
criticism of previous research (2 instances, one due to an error in interpretation),
passive form (3 cases), and presence in a subordinate or non-finite clause (3 cas-
es), or a combination of these. However, these alternative verbs have other uses
and functions so that care should be taken not to overuse them or to apply them
in inappropriate contexts (Williams 2007, forthcoming). There are Spanish verbs
within the semantic scope of those under study that tend to be underrepresented
because they have no direct equivalent in English or because they do not imme-
diately come into the translators mind in some contexts. For the presentation of
claims and findings with slight downtoning, both comprobar and constatar (show,
find, verify or confirm among other meanings) are used agentively in the first-
person or with named researchers as subject. For a statement of results in a com-
pletely neutral tone, either for modesty or to avoid conflict, obtener (obtain) is
often suitable for self-citation and referir (report) and comunicar (communicate,
often report) for citations of other research.
(22) However, such effects, at least for known risk factors, would be modest
compared with the large risks we showed.
> Sin embargo, dichos efectos, como mnimo, para los factores conocidos de
riesgo, seran modestos comparados con los importantes riesgos que hemos
mostrado. [*demostrado/comprobado/constatado].
In (22), if the agentive force of demostrado as a verb of communicative persua-
sion is felt to be too strong, either of the other two perception verbs could be used
rather than mostrado, which as we have seen appears to conflict with agentive pat-
terns. For even greater neutrality, los importantes riesgos obtenidos en este trabajo
(the large risks obtained in this study) would be suitable.
62 Ian A. Williams

(23) The rate of rim-area loss of 2.1%/y in eyes with an initial field defect and
1.7%/y in the contralateral eye is comparable with the rate recently shown by
Airaksinen et al.25
> El porcentaje de prdida del rea del borde papilar del 2,1%/ao en ojos
con un defecto inicial del campo y del 1,7%/ao en el ojo contralateral es
comparable con el porcentaje recientemente mostrado por Airaksinen et al25.

In (23), collocation of mostrado with porcentaje seems unnatural, and shown could
be rendered by constatado/ referido/ comunicado recientemente por Airaksinen y
cols., since any of these verbs collocates better with numerical expressions.
(24) A phagocytic overload was shown to produce cell injury and death by
Shirato et al.,27 who reported necrosis in trabecular cells that had ingested a
large number of plastic beads.
> Shirato et al.27, quienes informaron de necrosis en las clulas trabeculares
que haban ingerido un gran nmero de bolitas de plstico, sealaron que una
sobrecarga fagoctica produce lesin y muerte celulares.
* Shirato y cols27 han demostrado que una sobrecarga fagoctica produce lesin
y muerte celulares, al observar necrosis en las clulas trabeculares que haban
ingerido numerosas bolitas de plstico.

In example (24), thematization of A phagocytic overload is justified in English as


a linear link with the phagocytic process in the previous rheme, and this leads to
use of the hybrid form of show with the authors expressed as the passive agent, to
which the non-defining relative can be appended. In Spanish, indefinite concepts
are not usually felicitous as themes and there is a strong preference for the active
over the passive; the translator is, therefore, warranted in bringing Shirato et al.
forward as subject and theme, but the change in the position of the relative clause
is questionable, as is the choice of sealar for demonstrate. The proposed version
maintains the authors as subject but uses the canonical agentive demostrar in the
present perfect, which indicates relevance of the citation rather than remoteness
(Williams 2005), and restores the order of the two clauses by means of an infinitive
clause with the neutral observation verb replacing the neutral reporting verb to
provide the support for the claim. Interestingly, the range of meaning of compro-
bar would allow replacement of either demostrar or observar, with downtoning in
the former and upgrading in the latter case.

4.2 Translation in the Techniques category

For Techniques, in addition to the 29 instances translated by mostrar and dem-


ostrar in some form (Table8), the other contexts were rendered by poner de mani-
fiesto (make clear) (4 cases) although this verb did not appear in SNLC with
Show, demonstrate, mostrar and demostrar in medical articles 63

this function4 by evidenciar (evidence), observar (observe) and revelar (re-


veal) (twice each), ofrecer (offer) and lograr una visualizacin (achieve visualisa-
tion) (once each), with two cases of non-verbal translation.
Two instances with demostrar changed category. One changed to the Evi-
dence category with the addition of the noun resultados: (our IMC labeling []
has demonstrated striking differences between XD and HX); nuestros resultados
de MIC [] demuestran diferencias importantes entre el XD y la HX. The other
became passive with the alternative adjunct plus VS syntactic structure.
From the TL viewpoint, show and demonstrate also provide 67% (27 of 40) of
the Techniques category of mostrar and demostrar. The sources of the 13 tokens
of mostrar and demostrar not deriving from show and demonstrate were reveal (9
cases), disclose, display, prove and yield (once each). In spite of this use of mostrar
and demostrar for reveal, revelar was still overrepresented in STLC both for this
function and overall: 14 of 21 instances collocated with techniques as pseudoagent
subject, compared with only 1 of 4 in SNLC. Further gains could, therefore, be
made from revelar, since the Techniques category was slightly underrepresented
in STLC, epecially for demostrar.
In the Results section, whereas the distribution of mostrar and demostrar was
similar (10 tokens each) in SNLC, the former (16 tokens) exceeded the latter (5
tokens) in STLC, perhaps indicating a reluctance to use demostrar in this section
under the influence of the ESLC distribution of show and demonstrate (21 and 2
tokens, respectively). In contrast, Spanish writers may avoid mostrar in a section in
which the noun muestra/s (sample/s) is likely to appear as well as other functions
of the verb mostrar, particularly for Characteristics and metatextual reference to
tables and figures:
(25) The paired t test showed significantly less haze after reoperation at month
3 (P <05) and month 6 (P <01). [] Figure2 demonstrates a typical
evolution of scar formation.
> La prueba t para muestras pareadas muestra una opacidad
significativamente menor despus de la segunda intervencin, tanto en el tercer
mes (p <0,05) como en el sexto (p <0,01).[] La figura 2 muestra la evolucin
tpica de la formacin de una cicatriz.
* La prueba t para muestras relacionadas demuestra una opacidad
significativamente menor despus de la segunda intervencin, tanto en el
tercer mes (p <0,05) como en el sexto (p <0,01).[] En la figura 2 se refleja la
evolucin tpica de la formacin de una cicatriz.

In (25) three very similar lexical forms appear in close proximity in the translation
(one sentence has been omitted). Since the noun form constitutes a term and is
unavoidable, adjustments must be made in the verbs. The proposed version has
64 Ian A. Williams

demostrar rather than mostrar for the technique and incorporates changes both in
the syntax and in the lexical verb in the metatextual pointer.

4.3 Translation in the Evidence category

In the category Evidence, translators resorted to verbs other than mostrar or dem-
ostrar in only 9 contexts (Table8): evidenciar (evidence) (3 cases), indicar (indi-
cate) (3), poner de manifiesto (make clear) (2), and encontrar (find) (once). Thus,
translation of this function shows great consistency, with mostrar and demostrar
covering 88% of the total.
From the TL viewpoint, the two English verbs are the sources of 74 of the 83
tokens of mostrar and demostrar (89%). The other 9 sources were report (2 tokens),
establish, illustrate, observe, prove, provide (good) evidence, and reveal (once each),
with one equative sentence expressing a claim: (The main finding of this study is
that []); El hallazgo fundamental de este estudio muestra que [..], where mostrar
provides a stronger link than the copula.
Despite the great regularity in the translation of this function, the quantative
analysis indicates overrepresentation of both Spanish verbs, so that once again con-
statar and especially comprobar can be suggested as occasional alternatives, par-
ticularly in contexts in which mostrar and demostrar appear in close proximity.
(26) There have been several recent studies showing corticosteroids to be
efficacious in the management of acute asthma, but none demonstrated the
efficacy of oral therapy in the ED and none showed a corticosteroid benefit
in the setting of frequent 2-agonist therapy.
> Diversos estudios recientes han demostrado la eficacia de los corticoides en
el tratamiento del asma aguda, aunque ninguno de ellos demostr la eficacia
del tratamiento por va oral en el departamento de urgencias, y ninguno
demostr un beneficio de los corticoides en combinacin con inhalaciones
frecuentes de betadrenrgicos.
* Aunque diversos estudios recientes han demostrado la eficacia de los
corticoides en el tratamiento del asma aguda, ninguno de ellos ha demostrado
la eficacia de su administracin oral en el servicio de urgencias, y en ninguno
se ha comprobado un beneficio de los corticoides en combinacin con
inhalaciones frecuentes de betadrenrgicos.

In (26) two instances of show and one of demonstrate give rise to three tokens of
demostrar in a single complex sentence. The proposed version reduces the repeti-
tion by using comprobar with a change in syntax, since this verb, like many other
perception and communication verbs (Williams 2005),5 did not occur with an
evidencial pseudoagent subject in SNLC.
Show, demonstrate, mostrar and demostrar in medical articles 65

4.4 Translation in the Metatextual category

In the Metatextual category, of the 66 tokens of show, 39 were translated by


mostrar, and one involving a that-clause interpreting the evidence displayed in
a figure was correctly rendered by demostrar. Other TL verbs used were exponer
(show) (9 times), presentar (present) (5), indicar (indicate) (3), describir (de-
scribe) (2), detallar (detail) (2), observar (observe) (2), incluir (include) and
recoger (gather) (once each), with one case of non-verbal translation. The single
token of demonstrate with this function was translated by mostrar, as shown in
example (25). The remaining four sources of mostrar were display, give, indicate
and outline. If the minimal contribution of demonstrate and demostrar is ignored,
mostrar accounts for almost 60% of the instances of show while, from the TL view-
point, show contributes almost 90% of the 44 tokens of mostrar (Table4). Despite
the great regularity of this translation pairing, it leads, as shown in the quantitative
analysis and contextual study, to a distorted TL text lexical profile for this metatex-
tual function (Table5). This distortion is compounded by the transfer to Spanish
of the subject + active verb (SV) syntagm, which is not a natural choice according
to the statistical analysis. Of 26 occurrences of metatextual show in the active, 20
took the equivalent form with mostrar, one became active ilustrar, and only 5 in-
stances were recast in the canonical structure of adjunct + reflexive passive + post-
posed subject, each incidentally with a different verb, and only one with mostrar.
This minority translation strategy is to be recommended. Thus, SV syntax should
be avoided, and the discourse should be checked to decide how the phenomenon
and reference to table or figure should be distributed between theme and rheme.
Lexically, a variety of verbs should be employed, particularly reflejar for both ta-
bles and figures, recoger for tables, and representar for figures; in contrast, mostrar
and presentar should be used with care for metatextual reference because of their
multiple functions. In addition, in the case of the present-tense forms of mostrar,
there may be a clash with the noun muestra/s (sample/s), which is a common item
in many research reports.
(27) Table1 shows the demographic data of the patients and the relevant
conditions for which they were receiving treatment before the index
infarction.
Figure1 shows the findings for all-cause mortality.
Table2 shows all-cause mortality (primary endpoint) and secondary events
validated by the outcomes subcommittee (20).
The withdrawal rates from study drug, excluding patients who had died, are
shown in figure 2.
> En la tabla 1 se presentan los datos demogrficos de los pacientes y los
trastornos relevantes por los que reciban tratamiento antes del infarto inicial.
66 Ian A. Williams

En la figura 1 se recogen los resultados en cuanto a la mortalidad por el


conjunto de todas las causas.
En la tabla 2 se indica la mortalidad por todas las causas (variable de
valoracin final primaria) y los episodios secundarios validados por el
subcomit de resultados (20).
En la figura 2 se presentan los porcentajes de suspensin prematura del
frmaco en estudio, excluyendo a los pacientes fallecidos.

Example (27) includes four metatextual pointers from the Results section of an SL
text in which show has been used, three times in the active and once in the passive.
The translator has chosen the canonical adjunct structure in the four sequences
and has avoided use of mostrar in all cases. The example, therefore, illustrates a
number of the principles proposed here on the basis of the quantitative results.

4.5 Translation in the Characteristics category

In the Characteristics category, the data for human and inanimate subjects, as
indicated in Table8, will be dealt with separately. Apart from the 9 tokens of show
translated by mostrar, translators used presentar (present) (13 times), observar
(observe) (twice), and cumplir (fulfil) (once, a collocation with criterios (crite-
ria)), and one instance of non-verbal expression. Of the 3 tokens of demonstrate,
one became mostrar, one was translated by passive demostrar (see example (30)
below), and the third was recast with observar. Thus, mostrar and demostrar to-
gether accounted for only 38% of the characterisation of human subjects, fewer
than presentar (45%). From the TL viewpoint, in addition to the 9 instances deriv-
ing from show, other sources of mostrar were have (11 tokens), exhibit (2), demon-
strate and present with (once each). Therefore, show and demonstrate contributed
only 10 contexts of the total of 24 (42%), fewer again than the main source have
(46%).
This variability in translation behaviour may reflect uncertainty on the part of
the translator as to the collocatability of certain lexical items with mostrar in ad-
dition to other stylistic concerns such as excessive repetition of the more frequent
verbs expressing the characterisation function, presentar (present) and tener
(have). Nevertheless, the lexical choices usually reflect the main criteria or se-
mantic affinities identified in Section3.3: phenomena made visible by microscopy
or other special techniques; abstract entities obtained by laboratory or statistical
analysis. Translations of show by verbs other than mostrar often involved the more
obvious manifestations: e.g., detrusor instability, overlapping features, left ventricu-
lar enlargement, episode of desaturation, apneic pauses, which all combined with
presentar. In fact, choice of this verb is a safe option for both complement types
Show, demonstrate, mostrar and demostrar in medical articles 67

with a human subject, since it does not give rise to the ambiguity observed with
mostrar between the controlling Agent role and the less dynamic Possessor role.
For the contexts in which other SL verbs were translated by mostrar, again
most of the direct object collocates met the criteria: visible phenomena anti-
cuerpos (antibodies), depsitos de IgG (IgG deposits), inmunofluorescencia di-
recta positiva (positive direct immunofluorescence), neovascularizacin coroidea
(choroidal neovascularisation); abstractions inmunosupresin clnica (clini-
cal immunosuppression), caractersticas (characteristics), hallazgos (findings),
fracasos (failures), disminucin (decrease), refraccin (refraction). Interest-
ingly, general lexical items expressing clinical features and findings collocate with
mostrar, whereas specific items may not. However, in some contexts, the colloca-
tions were not entirely felicitous:
(28) Furthermore, OGRs [organ graft recipients] often presented with multiple
SCCs [squamous cell carcinomas], whereas all control SCCs were single
lesions.
> Adems, los PT [pacientes trasplantados] suelen mostrar [* presentar]
mltiples SCC, mientras que en el grupo control todos ellos fueron lesiones
nicas.

In a similar context in a Spanish article, presentar was combined with lesiones,


which were also in fact tumours.
(29) En nuestra serie, dos pacientes presentaron cuatro y seis lesiones []
(In our series, two patients presented four and six lesions [])

In one dermatology text, the translator did not use the verb tener for characterisa-
tion, and this led to four contexts in which outwardly visible manifestations (eye
problems and skin lesions) seemed unnatural and did not conform to the criteria:
(30) In that study, all eight patients who demonstrated positive serologic
reactivity to BP Ags had skin lesions.19
> En dicho estudio, los 8 pacientes en los que se demostr una reactividad
serolgica positiva frente a antgenos BP mostraron lesiones cutneas.19
* En dicho estudio, los 8 pacientes que mostraron una reactividad serolgica
positiva frente a antgenos BP tenan/ presentaban lesiones cutneas.19

In (30), whereas the characterisation with demonstrate is adequately dealt with by


use of the middle se-type passive, the case with mostrar is not successful mainly
because of the past-tense form, which implies an event and, therefore, agentiv-
ity. Interestingly, as the proposed version shows, mostrar could be used instead of
demostrar for the dynamic internal biological process, but the descriptive imper-
fect of tener or presentar is indicated in the second context. Indeed, as a general
68 Ian A. Williams

rule, it would seem that the more likely it is that specific technology or special
investigations are required to uncover the phenomenon, the more likely mostrar
will be suitable in the context. Conversely, the more obvious the evidence, the less
likely this will be the case, since the human subject of mostrar could be taken as
volitional Agent rather than fulfilling the role of Possessor. In such cases, presen-
tar or tener are suggested as safe alternatives.
A further anomalous case involved a statistical abstraction. Here, the SL text,
which itself is not entirely natural, may have misled the translator:
(31) These three patient subsets and the BP control group showed similar average
ages of onset.
> Estos tres subgrupos de pacientes y el grupo control de BP mostraron una
edad media de inicio de la sintomatologa similar.
* Estos tres subgrupos y el grupo de referencia tenan una media de edad similar
al inicio de los sntomas.

In SNLC, edad (age) almost invariably collocated with the verb tener (have).
The translation of the Characteristics function with inanimate entities was
far more consistent. In addition to the 35 tokens that became mostrar or dem-
ostrar (Table8), 14 instances were rendered by other verbs: presentar (present)
(5 instances), and detectar (detect), observar (observe), obtener (obtain), resul-
tar (prove to be), and tener (have) (once each), with 4 instances of non-verbal
translation. From the TL viewpoint, show and demonstrate contribute 35 of 44
contexts (79%) in which mostrar and demostrar appear with this type of subject.
The remaining 9 sources for mostrar were have (3 times), display (twice), with 4
instances added through reformulation.
It should be noted, however, that use of both demonstrate and demostrar is rare
and appears suitable only to describe the active properties of drugs and therapeu-
tic procedures (e.g., inhibicin (inhibition) or eficacia (efficacy)) or biological
behaviour, as shown in (32):
(32) Phytohemagglutinin blasts generated from 14 patients with CTCL
demonstrated a significantly reduced proliferation (P <05) in the presence
of all concentrations of IL-2 compared with 15 control subjects.
> Los blastos inducidos por PHA generados a partir de la sangre de 14
pacientes con LCCT demostraron una proliferacin significativamente
reducida (P <0,05) en presencia de todas las concentraciones de IL-2, en
comparacin con 15 controles.
Show, demonstrate, mostrar and demostrar in medical articles 69

4.6 Translation in the syntactic categories


Despite the systemic differences illustrated in Section2.2. for the syntactic cat-
egories, the quantitative analysis showed that translators were largely successful
in making the changes required between languages. In SNLC, the passive imply-
ing agentive evidentiality is expressed with demostrar. Accordingly, 7 of 11 tokens
of passive show and 16 of 19 instances of demonstrate became demostrar in the
true or se-type passive. Another 6 tokens (3 for each verb) were rendered by more
neutral reporting or perception verbs: comprobar (verify) (twice), and describir
(describe), encontrar (find), evidenciar (evidence) and observar (observe). The
single instance of show transferred as mostrar would have been better expressed
with demostrar: (the effectiveness of acute and long term blockade has been
shown) Se ha [de]mostrado la eficacia del bloqueo beta. An additional instance
of passive mostrar with agentive implications derived from display and described
a procedure in the Methods section, a context in which use of mostrar is unusual,
and perhaps could be better rendered by a verb like exponer, which expresses only
the agentive display-type show, as illustrated in (33):
(33) At the end of the study each participants radiographs were displayed
simultaneously in chronological order for blind assessment of vertebrae T4
to L5.
> Al trmino del estudio, se mostraron las radiografas de cada participante de
una manera simultnea, en orden cronolgico, para una valoracin ciega de
las vrtebras T4 a L5.
* Al trmino del estudio, las radiografas de cada participante fueron expuestas
simultneamente y en orden cronolgico para la valoracin enmascarada de
las vrtebras de T4 a L5.
Passive demostrar also received 9 instances proceeding from other structures with
show (Hybrids 7 and Evidence 2) and 4 from other active categories of demon-
strate, namely Authors, Evidence, Techniques, and Characteristics, as in exam-
ple (30). Gains also came from 12 other sources: prove (3 times), establish (twice),
describe, document, found to have, study and substantiate, with two recast from
other structures: is demonstrable and has not been the case. Thus, show and dem-
onstrate contribute 37 of the 50 (74%) passive contexts of mostrar and demostrar,
with an overwhelming predominance of the latter (Table4).
Of the 19 show Hybrids, one was reformulated in the active Evidence cat-
egory, another was translated by observar, and two were expressed non-verbally.
However, as just mentioned, a major destination (7 cases) is passive demostrar
in the se form with VS inversion and the subject expressed as a clause. This is a
particularly suitable solution when the subject of the English hybrid is new to the
discourse or structurally complex, as in (34):
70 Ian A. Williams

(34) Besides complex formation, increased production of elastase, which


overwhelms the inhibitor system, as well as oxidative inactivation of
1-proteinase inhibitor have been shown to contribute to the protease-
antiprotease imbalance in some neonates with RDS.1,10,11
> Adems de la formacin de complejo, se ha demostrado que tanto la mayor
produccin de elastasa, que desborda el sistema inhibidor, como la inactivacin
oxidativa del 1-PI contribuyen al desequilibrio proteasa-antiproteasa en
algunos recin nacidos afectados de un SDR1,10,11.

Another 7 instances were translated by the Reflexive demostrar. However, it


should be remembered that the Spanish authors showed a slight preference for the
forms with mostrar, which can be used where English has shown to be. Thus, in
(35), ha demostrado ser could be replaced by se ha mostrado como:
(35) In this study, photodynamic therapy has been shown to be an effective
therapeutic modality for the treatment of skin cancer.
> En este estudio, la terapia fotodinmica ha demostrado ser una modalidad
teraputica efectiva para el tratamiento de cncer cutneo.

The final instance was translated by an unattested and apparently anomalous form
of mostrar + infinitive:
(36) Of the 74 patients who became urease negative, one was subsequently shown
to be positive for H pylori by culture.
> De los 74 pacientes que pasaron a ser ureasa negativos, uno mostr ser
positivo posteriormente a H pylori a travs del cultivo.
* De los 74 pacientes que pasaron a ser negativos en la prueba de la ureasa, uno
posteriormente result ser positivo para H pylori mediante cultivo.

The proposed version uses the verb resultar ser (prove to be), which avoids any
agentive overtones of a human subject plus Reflexive mostrar or demostrar.
Interestingly, the remaining 6 cases of Reflexive mostrar in the TL texts derived
from recast relational sentences with strengthening of the copula with mostrarse.
For demostrar, in addition to the 7 tokens from the show hybrids, another 3 oc-
currences proceeded from other categories with this verb (Characteristics 2 and
Evidence 1), the remaining sources being prove (3 times) and one instance of
reinforcement of the copula. Thus, the contribution of show to this function of
demostrar was 71%.
Show, demonstrate, mostrar and demostrar in medical articles 71

4.7 Recommendations for English-Spanish translation

On the basis of the results of this study the following guidelines can be proposed
for English-Spanish translation of statements containing show and demonstrate in
the different environments.
1. In the Authors category, both show and demonstrate can be rendered by dem-
ostrar, with mostrar only used for show meaning display. Other local options
are comprobar and constatar to avoid excessive repetition, or if slight down-
toning is appropriate.
2. For Techniques, both mostrar and demostrar are suitable for the two English
verbs, equal frequency being the aim and demostrar especially apt in the Results
section, where mostrar has other functions (Metatextual and Characteris-
tics) and present-tense forms can clash with the noun muestra/s (sample/s).
3. In the Evidence category, demostrar is to be preferred to mostrar in a ratio of
2:1. Excesses of the two verbs can be avoided by resorting to other options:
poner de manifiesto, or comprobar and constatar with a change to adjunct + VS
syntax.
4. For Metatextual reference, mostrar is the choice, but the aim should be lexical
variety, through use of other verbs such as exponer, reflejar, recoger and repre-
sentar. Subject-verb syntax should be avoided and the context checked for the
optimal thematic choice between phenomenon and extratextual item.
5. For Characteristics, mostrar is again the predominant of the two Spanish
verbs. Attention should be paid to the subject and direct object collocates:
lexical items expressing outwardly evident manifestations should be avoided
with human subjects, these collocating better with presentar and tener. With
inanimate subjects, mostrar is usually appropriate, but demostrar is an occa-
sional alternative to convey biological behaviour and active properties in some
contexts.
6. For Passives, demostrar is first choice. The se-type should not be overused: the
ser + participle or the pseudo-form with quedar are alternative options in the
present perfect to express claims and RPRs, while estar + participle is appro-
priate for consensus statements.
7. English Hybrids with passive show + infinitive can be expressed by demostrar
with VS syntax and postposed clausal subject. Mostrarse translates Hybrids
with to be, whereas demostrar is possible with all infinitives, the frequency of
the two verbs for this function being fairly even. The inagentive resultar + in-
finitive is an apt alternative for these Reflexive structures in some contexts.
72 Ian A. Williams

5. Conclusion

This study confirms the register-dependent distribution of functions for show


(Biber et al. 1999) and extends this finding to mostrar and demostrar as seen in
this extensive corpus of medical RAs. The analyses show that the considerable
regularity of the translation correspondence between the English SL verbs show
and demonstrate and the two Spanish verbs (268 of 368 tokens, 73%) masks varia-
tion in lexical selection across the environments defined in the study.
The almost exclusive choice of mostrar in the Metatextual and Characteris-
tics categories corresponds to agentive put on view and inagentive display, re-
spectively. In the first case, agentivity is left implicit since the preferred syntax is
with the se-type middle structure, and occasionally with a metatextual pseudo-
agent in the active. For Characteristics, there is a weak Possessor (human or
inanimate) which in this register appears incompatible with a stronger agentive
meaning, since the structure with the indirect object Receiver was not attested.
This inagentive use also occurs in the Reflexive middle structure with mostrarse.
In contrast, the almost exclusive choice of demostrar corresponds to the mean-
ing make something known or clear, and is agentive, explicitly in the Authors
category and implicitly in the Passive category. The Reflexive use of demostrar
shares the display meaning of mostrarse, but is more dynamic with the Possessor
exhibiting some active quality or behaviour. In the Techniques and Evidence
categories, both mostrar and demostrar correspond to the meaning of reveal or
disclose, with the procedure and data source acting as the pseudoagent subject
and human agency remaining implicit.
The results of the quantitative and contextual analyses show that translation
studies based on specialised corpora contribute usefully to the semantic analysis
of polysemous lexical items involved in complex cross-linguistic patterns, and can
provide strong empirical evidence for the formulation of recommendations for
translation practice.

Notes

1. In the examples, the relevant verbs are shown in bold and other features of note are under-
lined. Examples from ESLC are followed when appropriate by the corresponding translation,
which is signalled by the symbol (>), and sometimes by a revised version proposed on the basis
of the results of this study and indicated by an asterisk (*). Examples from SNLC are followed by
a gloss or a natural translation (provided by the author) as appropriate. It should be noted that
all superscript numbers in the examples are bibliographic references in the cited text, and have
no other significance within the present paper.
Show, demonstrate, mostrar and demostrar in medical articles 73

2. The Others category includes for show 8 instances of parenthetical use, e.g. data/illustration
not shown, and one instance of infinitival use easier to demonstrate, which could not be classi-
fied into any of the categories. Similarly, SNLC had 8 instances that could not be classified: 5
instances of adjectival use e.g. diabticos con complicaciones vasculares demostradas (diabetics
with proven vascular complications) and three infinitives. Finally, the TL texts had 5 instances
of the parenthetical function (no se muestran los datos) and one adjectival use for mostrar, and
3 instances of adjectival use and one infinitive for demostrar. Together, they represent less than
3% of the total sample and will not be considered further here.

3. It should be remembered that the few cases of the passive with a human agent were included
in the Authors category.

4. Poner de manifiesto appeared 10 times in SNLC: Evidence (5 tokens), Passives (3), the re-
maining 2 being adjectival uses with the data source and technique in an adjunct.

5. The study covered the Discussion section only, the most frequent verb with an evidencial
pseudoagent, in what I term anthropomorphic metonymies (research for researchers), was dem-
ostrar (12 tokens); other communication or perception verbs were confirmar (5), sugerir (3),
poner de manifiesto (2), encontrar (2), concluir, indicar, referir and revelar (once each). Pseudo-
agent subjects are avoided through the preferred syntax of Adjunct + VS inversion.

References

Biber, D., Johansson, S., Leech, G., Conrad, S. & Finegan, E. (1999). Longman Grammar of Spo-
ken and Written English. London: Longman.
Brett, P. (1994). A genre analysis of the results section of sociology articles. English for Specific
Purposes, 13 (1), 4759.
Collins COBUILD English Language Dictionary (1987). Sinclair, J., Hanks, P., Fox, G., Moon, R.
& Stock, P. (Eds.). London/Glasgow: William Collins Sons & Co. Ltd.
Diccionario de la Real Academia Espaola (22 edicin) (2001). Available at: http://www.rae.es
(last accessed October 2007)
Dudley-Evans, A. (1994). Genre analysis: an approach to text analysis for ESP. In M. Coulthard
(Ed.), Advances in Written Text Analysis (pp. 219228). London: Routledge.
Laviosa, S. (2002). Corpus-based Translation Studies:Theory, Findings, Applications. Amsterdam/
New York: Rodopi.
Nava, F. & Maldonado, R. (2004). Basic voice patterns in Tarascan (Porhepecha). In M. Achard
& S. Kemmer (Eds.), Language Culture and Mind (pp. 461478). Standford: CSLI Publica-
tions.
Neubert, A. & Shreve, G. (1992). Translation as Text. Kent, Ohio/London: The Kent State Uni-
versity Press.
Scott, M. (1998). WordSmith Tools Manual. Oxford: Oxford University Press.
Tognini-Bonelli, E. (2001). Corpus Linguistics at Work. Amsterdam: John Benjamins
Williams, I. A. (forthcoming). A corpus-based study of Spanish translations of the verb report
in biomedical research articles. To appear in Meta.
74 Ian A. Williams

Williams, I. A. (2007). A corpus-based study of the verb observar in English-Spanish transla-


tions of biomedical research articles. Target, 19 (1), 85103.
Williams, I. A. (2006). Towards a target-oriented model for quantitative contrastive analysis
in translation studies: An exploratory study of theme-rheme structure in Spanish-English
biomedical research articles. Languages in Contrast, 6 (1), 145.
Williams, I. A. (2005). Thematic items referring to research and researchers in the Discussion
section of Spanish Biomedical articles and English-Spanish translations. Babel, 51 (2),
124160.
Williams, I. A. (1999). Results sections of medical research articles: Analysis of rhetorical cat-
egories for pedagogical purposes. English for Specific Purposes, 18 (4), 347366.
Williams, I. A. (1996a). A translators reference needs: Dictionaries or parallel texts? Target, 8
(2), 275299.
Williams, I. A. (1996b). A contextual study of lexical verbs in two types of medical research
reports: Clinical and experimental. English for Specific Purposes, 15 (3), 175197.

Authors address
Ian A. Williams
Universidad de Cantabria, Departamento de Filologa,
Avenida de los Castros s/n, 39005 Santander, Spain
williams@unican.es

You might also like