Iberica 13
Ibérica 43 (2022): 77-102
ISSN: 1139-7241 / e-ISSN: 2340-2784
Abstract
A corporate social responsibility (CSR) report is a form of information
disclosure about environmental, health, and human rights issues. It serves as an
essential bridge for communication between enterprises and stakeholders.
Enterprises can use vague language (VL) to strategically communicate with the
public, establish a positive corporate image, and legitimise their activities. In
recent years, many scholars have studied the usage and communicative function
of VL in general conversational discourse, yet VL in business discourse has not
received much attention.
In this study, 30 social responsibility reports of 10 world-renowned cosmetic
enterprises between 2015–2019 were collected to build a corpus. With the help
of the corpus software AntConc, this paper explores VL, specifically in CSR
reports, using a corpus-assisted approach. Further, the quantitative findings are
complemented by the contextualised interpretation of the communicative
functions of VL in CSR reports. The major findings of this research are as
follows. There are four types of VL used in CSR texts: quantity, degree, time, and
softening stance-taking. In addition, the use of VL can help achieve several
communicative purposes of CSR report drafting (i.e., provide an appropriate
amount of information, enhance persuasion, and self-protection). Some
practical suggestions are discussed for improving the awareness of the strategic
use of vagueness in CSR reports.
Keywords: vague language, corporate social responsibility report, corpus-
assisted discourse analysis, cosmetics industry, business communication.
A corpus-assisted study of vague
language in corporate responsibility
reports of the cosmetics industry
Bixi Jin
Beihang University (China)
bixijin@outlook.com
77
Ibérica 43 (2022): 77-102
BIxI JIn
Resumen
Estudio asistido por corpus del lenguaje vago en los informes de responsabilidad
social corporativa de la industria cosmética
El informe de responsabilidad social corporativa (RSC) constituye una forma de
divulgar información sobre asuntos relacionados con el medio ambiente, la salud
y los derechos humanos. Funciona como un puente esencial que posibilita la
comunicación entre las empresas y los accionistas. Las empresas pueden emplear
un lenguaje vago (LV) de manera estratégica para comunicarse con sus
destinatarios, establecer una imagen corporativa positiva y legitimar sus
actividades. En los últimos años, muchos investigadores han estudiado el uso y
la función comunicativa del LV en el discurso conversacional general. Sin
embargo, el LV apenas ha sido analizado en el discurso empresarial.
Para el presente trabajo se ha compilado un corpus de 30 informes de
responsabilidad social de 10 empresas mundialmente conocidas de la industria
cosmética publicados durante los años 2015-2019. Con la ayuda del gestor de
corpus AntConc se ha explorado el LV en los RSC desde una perspectiva asistida
por corpus. Los resultados cuantitativos se han complementado con una
interpretación contextualizada de las funciones comunicativas desempeñadas
por el LV en los informes de RSC. Entre los hallazgos obtenidos, cabe destacar
que se han identificado cuatro tipos de LV en los textos analizados: de cantidad,
de grado, de tiempo y de toma de postura de carácter atenuante. Además, este
trabajo ha evidenciado que el uso del LV puede ayudar a alcanzar diferentes
propósitos comunicativos en los informes de RSC: proporcionar una cantidad
adecuada de información, mejorar la persuasión y autoprotegerse. También se
ofrecen algunas sugerencias prácticas para que se tenga una mayor conciencia del
papel estratégico de la vaguedad en los informes de RSC.
Palabras clave: lenguaje vago, informe de responsabilidad social
corporativa, análisis de discurso asistido por corpus, industria cosmética,
comunicación empresarial.
1. Introduction
Extensive research has been conducted to provide various definitions of
corporate social responsibility (CSR) and related notions (e.g., sustainability
reports), and the evolving conceptualizations of CSR over the decade
reveal its inherent intricacies (Maignan, Ferrell, & Ferrell, 2005). A CSR
report is officially defined as “the responsibility of enterprises for their
impact on society” and “a process to integrate social, environmental,
78
ethical, human rights and consumer concerns into their business
operations and core strategy in close collaboration with their stakeholders”
(European Commission, 2011: 6). Though it is becoming a generalised
practice to include a social responsibility section in annual reports, as the
L’oreal report (2018) indicates, this non-financial reporting section
achieves the communicative purpose of presenting a clear picture of a
company’s efforts and commitments in environmental sustainability, ethics,
and corporate governance issues, seeking to obtain transparency and
legitimacy (ditlevsen, nielsen & Thomsen, 2013). Further, it is a
management strategy to construct a favorable and trustworthy corporate
image (van Riel & Fombrun, 2007; Kushal, 2011; Fuoli & hart, 2018),
targeting a group of heterogeneous audiences, including “existing
shareholders, potential shareholders, creditors, investment and credit
analysts, the government and their agencies, employees and society at
large” (ditlevsen, 2012: 382). While CSRs tend to prioritise social
responsibility at large, informed by stakeholder theory, the incorporation
of stakeholder views and interests into CSRs has gained some prominence
in view of the unforeseen outcomes emerging from business activities
(Fordham & Robinson, 2018; Maignan et al., 2005).
As suggested by the well-known consulting company Ecovia Intelligence,
consumers in the cosmetics industry have become extremely sensitive to the
range of social and environmental initiatives mentioned in CSR reports.
growing consumer awareness prompts companies to be particularly
concerned to explain to the public such facts as basic ingredients in the
manufacturing process, animal testing, the certification of natural and
organic cosmetics, sustainable packaging, and even financial assistance to
disadvantaged communities. As a result, a company may organise CSR
reports with a range of linguistic means to demonstrate its social and
environmental initiatives. use of vague language (VL), recognised by its
communicative importance, is a drafting strategy in organizing specialised
discourse such as laws, policy documents, and business communication
(handford, 2010; Li, 2017; McCarthy, 2019; Scott di Carlo, 2015). The
strategic use of VL may also contribute to the delicate discursive negotiation
in CSR reporting. The possible application of VL in CSRs motivates the
current research.
Business contexts are appropriate places to observe the use of VL. on the
one hand, it may be tempting to think that companies are expected to
employ VL to emphasise or amplify their positive efforts in maintaining
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 79
positive corporate images and legitimizing corporate actions; on the other
hand, VL may ensure the desired level of flexibility for companies to
maintain control over private information or account for unexpected future
contingencies of corporate activities (Choi & Triantis, 2010).
Therefore, this study aims to identify and explore the strategic use of VL in
CSR reports in the cosmetics industry. The following questions are
addressed: (1) What types of VL are prototypical of CSR reports in the
cosmetics industry? (2) Is the use of VL in CSR reports related to their
communicative purposes? If so, how?
The study seeks to make several contributions to the current literature on
corporate discourse. First, the findings enrich the current understanding of
VL in corporate discourse. The quantitative corpus linguistics methods and
the detailed qualitative discourse analysis of VL in the beauty industry CSR
reports elucidate the use of VL in the construction of CSR reports.
Second, from a methodological perspective, the study explores VL in CSR
reports based on a relatively large-scale specialised corpus, enabling the
particular language expressions in rich contexts to be identified and
categorised. Third, the study provides a detailed account of VL in CSR
reports and examines how the reports may strategically communicate with
heterogeneous audiences, construct a positive corporate image, and
legitimise CSR activities. given that there is no readily available
instructional space for VL, the results address the need for incorporating
VL into future business training programs and business English learning
materials.
2. Definition of vague language
Considerable discussion in recent years has focused on the definitions of
VL. While a philosophical approach has suggested that language is
inherently vague (Russell, 1997), this is not practical for rhetorical-
linguistic analysis. Research addressing the definition of VL has generated
considerable interest among applied linguists; however, the internal
configuration of this term remains unclear, and even the terms used to
refer to VL are themselves vague (Cotterill, 2007). VL has been interpreted
in several ways, and presented here are some often-cited definitions. For
instance, Channell (1994: 20) argued that the evaluation of VL can be
determined by the replacement of other words or expressions if they
BIxI JIn
Ibérica 43 (2022): 77-10280
appear to “render the same proposition” and is “purposely and
unabashedly vague”. Pinkal (1995:12-15) first differentiates between
“communicative indeterminacy” and “semantic indeterminacy”. Then, he
establishes that there are two types of communicative indeterminacy:
vagueness and ambiguity.1 A similar criterion of determining VL was
proposed, focusing on the possibility of making the utterance precise and
the indeterminacy arising from linguistic expressions. Ruzaitė (2007: 28),
focusing on VL forms and functions, argues that “vague language is a
natural, usually purposeful and multi-functional linguistic phenomenon
that involves imprecision and is employed for certain communicative
strategies.” Zhang (2018: 7) continues this notion by acknowledging that
VL can be negotiated to adapt to specific communicative purposes and
further specifies that VL is unspecified and context-sensitive.
Based on the abovementioned studies, it is not surprising that the unclear,
indefinite configuration of VL makes it difficult to identify VL in actual
practice, as many researchers may propose their own interpretations.
Therefore, given the intricate nature of identifying VL, it is not my aim to
unravel the nuances and minutiae of the various definitions but to be aware
of the vague expressions identified by previous studies. The working
definition of VL in this study emphasises its unspecified, elastic, negotiable,
and communication-sensitive nature (Channell, 1994; Ruzaitė, 2007; Zhang,
2018), which will guide the subsequent data analysis and interpretation of
CSR reports in corporate discourse.
3. Vague language used in discourse
VL is a pervasive phenomenon in everyday language. Increasing attention
has been paid to the usages and communicative functions of VL in general
conversational contexts (Evison, McCarthy & o’Keefee, 2007; Cheng &
o’Keeffe, 2015) and more specialised discourse domains, such as political
discourse (gruber, 1993), courtroom discourse (Cotterill, 2007), and legal
discourse (Anesa, 2007; Li, 2017). The purposeful use of VL in these
contexts has been well investigated and summarised. For instance, Anesa
(2007) adopted the term “weasel words” to characterise words such as good,
appropriate, and reasonable in contracts, as informed by Mellinkoff (1963:
21). Li (2017: 105) created a VL list based on prior studies and categorised
the VL used in large-scale legal documents into time, category, quantity, and
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 81
degree, and proposed four pragmatic functions: extending applicability,
providing flexibility, maintaining a balance between precision and over-
elaboration, and mitigating potential problems. Zhang (2018: 123) identified
the functions of some under four pragmatic dimensions in natural classroom
settings: inclusion of an appropriate amount of information, mitigation,
withholding information, and discourse management.
despite its importance in business contexts, limited research regarding VL
has been conducted in this field. Liu (2015) discussed the uses of VL in
written business communication across cultures, such as politeness, face-
protection, and goal achievement. Malyuga and McCarthy (2018) observed
that the pragmatic functions of vague category markers (e.g., and so on, or
whatever, and that) in spoken business communication was manifested by its
knowledge-sharing practices among different business parties to form group
solidarity (also see handford, 2010), or it was used as a face-protection
strategy to forestall unexpected circumstances. While VL usages and
communicative functions have been well-investigated in general
conversational and specialised domains, there has been scarce discussion
about VL in the business discourse, particularly regarding the VL usages and
functions in CSR discourse.
4. Previous discourse studies of CSR reporting
There is a growing literature concerning CSR discourse. The CSR report is
generally recognised as a hybrid genre combining informative and
promotional elements (Bondi, 2016). Existing literature, on the one hand,
has focused on the description of generic structures of CSR reports based
on Swales’s (1990) and Bhatia’s (2004) frameworks (e.g., Skulstad, 2008;
Bhatia, 2012; yu & Bondi, 2017). on the other hand, some studies have been
committed to examining the lexico-grammatical resources utilised in the
CSR report. Findings suggest that the linguistic resources companies
mobilise for promoting socially responsible actions are used to construct a
positive corporate image. Malavasi (2011) observed the frequency wordlists
of CSR texts from two companies and analysed the predominant language
patterns revealing pervasive corporate values. Bondi (2016) investigated the
keywords and phraseologies related to prediction or commitment
statements, which are used to legitimise corporate activities and enhance
impression management. Stance resources related to expressions of attitudes
BIxI JIn
Ibérica 43 (2022): 77-10282
and evaluations have been much discussed in this area of inquiry. Fuoli
(2012) addressed the appraisal resources informed by systemic functional
linguistics (SFL) in the CSR reports by IKEA and BP. he concluded that
they employed interpersonal resources to portray a trustworthy corporate
image. Later, Fuoli (2017) investigated the stance expressions proposed by
Biber et al. (1999) in European- and uS-based CSR reports, focusing on
their efforts to establish committed, caring, and honest corporate images.
Informed by legitimation theory, Lin (2021) examined the discursive
strategies adopted by uK and China CSR reports to legitimise negative news.
Socio-cultural factors accounting for linguistic variation were discussed,
which contribute to the improvement of crisis management strategies used
by large companies.
5. Methodology
5.1. Corpus-assisted discourse analysis
A corpus-assisted discourse analysis approach was adopted for this study.
The CAdS approach is characterised by its synergy of corpus-aided analysis
and detailed qualitative discourse analysis, i.e., the automatic detection of
linguistic patterns (e.g., frequency, keywords, and collocations) from the
naturally occurring texts aided by corpus-analytic tools, enhanced by
elaborated interpretations of quantitative patterns in relation to their wider
social contexts (Partington, duguid & Taylor, 2013). The procedure starts
from a quantitative analysis aided by corpus techniques such as word
frequency and keywords to identify an overall linguistic profile. As a research
project unfolds, the analysis become more qualitative and context-sensitive
involving substantial efforts of human interpretation and explanation, with
less reliance on corpus tools (Baker & McEnery, 2015).
5.2. Corpus construction
In this study, a self-compiled corpus of annual CSR reports was built and
employed, consisting of annual CSR reports selected from 10 well-
established international beauty groups (L’oRÉAL, Estee Lauder,
PRoCTER & gAMBLE, KoSÉ, Johnson & Johnson, Amorepacific
group, LVhM, henkel, CoLgATE, and Beiersdorf). It contained 30 texts
(three texts selected from each corporation at random) produced from
2015 to 2019, which amounted to 775,237 words. As the study adopted a
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 83
CAdS approach, the corpus under construction was not committed to
compiling large corpora of various genre types, seeking to “obtain an
overview of the greatest quantity and variety of discourse types possible”
(Partington et al., 2013: 11-12). Specialised corpora of this kind are
designed to explore “specific types of genres within specific contexts”
(Connor & upton, 2004: 7-8) that could not be achieved with corpora built
for general purposes. Working on a manageably sized corpus also allows
researchers to familiarise themselves with the context of the texts and thus
enhances the interpretation of corpus findings beyond simply reading
concordance lines, countering the criticism that corpus-based research is
restricted to somewhat bottom–up lexico-grammatical analysis (L.
Flowerdew, 2005).
Table 1. Description of the CSR corpus.
5.3. Analytical procedure
Antconc 3.5.8 (Anthony, 2019) was employed in processing the corpus data.
It is a corpus analysis toolkit for text analysis, including major functions such
as wordlists, keywords, and concordances. The software can demonstrate the
concordance lines of a searched word or multi-word sequences with the
contexts provided. The analysis was conducted in three steps. In the first
step, ten texts were randomly selected and read from the total of 30 texts in
the corpus. The expressions related to VL were marked, considering the VL
items that have been identified by prior studies (e.g., Channell, 1994; Anesa,
2007; Cutting, 2007; Cheng & o’Keeffe, 2015; Li, 2017; Malyuga &
McCarthy, 2018). The point of conducting a close reading is that some vague
expressions cannot be automatically identified by the text analysis software.
Through a detailed reading of the sampled CSR texts, the researcher must
generate an initial list of the vague items in the CSR reports. The VL
generated in the list generally fell into the categories of approximate
quantities, boosters, downtowners, and shields (see Zhang, 2011: 574).
In the second step, I used AntConc to generate the top 50 words from the
wordlist and the keyword list (see Appendix) of the self-built CSR cosmetics
BIxI JIn
Ibérica 43 (2022): 77-10284
Corpus Number of texts Number of tokens Number of
types
Average
words of CSR
CSR 30 775237 essay 20031 25841
I
I
corpus to examine whether words or related phrases associated with
vagueness in the corpus could be added to the identified VL list. This step
aimed to supplement the vague patterns in step with the vague expressions
that may not have appeared in the 10 sampled texts. After checking,
expressions such as ‘more’, ‘or *’ and ‘all’ were added to the list obtained from
the Step 1 .2
In the third step, AntConc was used to analyse all the VL items identified in
Steps 1 and 2 by searching them throughout the corpus. The analyses
address several issues: the frequency and semantic categories of vague
expressions, the contexts where VL occurs, and the communicative
functions for using them in CSR texts. Meanwhile, a further investigation
with a research assistant specialised in applied linguistics was conducted to
exclude lexical items that were obviously not intended for vague use.
Examples were provided to illustrate how VL is strategically used in CSR
reports and to analyse the communicative purposes that the corporations
intend to achieve.
The quantitative corpus analysis of the specialised corpus categorises the VL
in CSR reports into four main types: quantity, degree, time and softening stance-
taking. overall, these four categories show the strategic role performed by
VL expressions in promoting a trustworthy corporate image. Companies
employ these linguistic resources as discursive tactics to “address
shareholders’ information needs, but also project readiness and competence;
they show that they are in control of the situation and prepared for any
contingency” (Fuoli, 2017: 21). In the following sections, VL representing
each category is discussed in detail.
6. Results and discussion
6.1. Vague items associated with “quantity”
Compared with other conversational discourse, language in corporate
discourse is supposed to be more precise (McCarthy, 2019). It is somewhat
surprising to see that the vague items associated with quantity, such as more
than, many, various, and some used in CSR texts, are at the forefront of
frequency level (5.2 per thousand).3 The recurrence of vague quantity avoids
revealing a specific numerical value to define the exact quantity level. A
possible explanation for this phenomenon is that companies are trying to
balance against accuracy and over-explicitness regarding reported corporate
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 85
activities in case of alternative interpretations. Examples 1–5 in the following
were identified for this communicative purpose.
Example 1. Additionally, in the past three years alone, this group has helped
more than 400,000 American families receive a new recycling cart, totaling more
than $27 million worth of new infrastructure and has forged several state-wide
partnerships to drive best practice adoption. (PROCTER & GAMBLE,
2017)
Example 2. While the world is making tremendous progress to curb the course
of the global HIV epidemic, many challenges remain in the areas of prevention,
treatment, access and public education. (COLGATE, 2015)
Example 3. This has allowed us to build a portfolio of successful brands in
various global markets and to become the leading cosmetic company representing
Asian beauty and its wisdom. (Amorepacific Group, 2016)
Example 4. Through this ongoing effort, we no longer use or are on our way to
eliminating some of the ingredients currently raising consumer questions.
(COLGATE, 2015)
Example 5. In addition, we have been implementing several programs to ensure
that our employees can pursue both parenting and professional growth.
(Amorepacific Group, 2016)
In Example 1, the drafter used “more than + an approximate number” instead of
the specific number, which is an estimation of quantity. The VL in this
example enables the drafter to avoid providing accurate data but does not
affect the dissemination of sufficient CSR information at the same time, as
it is unnecessary for the target audience to know the specific figures. Another
possible explanation for the use of this structure is that the Procter &
gamble (P&g) company does not have specific numbers either. The
intentional vagueness may be used to make up for a lack of information and
data.
In Examples 2 and 3, several instances such as m an y challenges and in va r i o us
global markets were found to make exact challenges and markets unclear.
While these usages of VL decrease the precision of language, it can be
inferred that CoLgATE and the Amorepacific group tend to emphasise
the efforts they have invested for the prevention of AIdS. The strategic
employment of these linguistic devices allows drafters to summarise the
current situation of AIdS prevention and the company’s globalisation
BIxI JIn
Ibérica 43 (2022): 77-10286
process and maintain an optimum degree of elaboration on the corporate
action.
Example 4 involves the vague word some (e.g., some of the ingredients). In the
CoLgATE report, relevant ingredients of cosmetic products and the
specific amount of them were not listed. In this context, the drafter
intentionally used VL to withhold ingredient information closely related to
quality problems in their cosmetics products. Instead of mentioning specific
figures, VL was employed as a strategic linguistic tool to avoid disclosing
internal negative information that may threaten its legitimacy in socially
responsible actions and cause image crises.
In Example 5, VL provides a generality of meaning. S ev er a l programs
conveys a general meaning instead of a specific one. Amorepacific group
tends to convey that they are committed to helping their employees achieve
a work–family balance. “Several” suggests that they invested consistent
efforts in establishing programmes, regardless of the exact number.
Furthermore, an emerging frequency of the pattern “a + singular quantifier +
of” should be noted, such as a variety of, a range of, a series of, and a number of,
which were also noted in Arinas’s (2012) study that similar expressions can
provide a flexible interpretation of the property defined in patents. here are
some examples of the use of a range of and a series of:
Example 6. Throughout the year, Amorepacific Group undertakes a range of
activities to effectively respond to various risks identified through our risk
mitigation and management process. (Amorepacific Group,2016)
Example 7. At our Vision Care facility in Limerick, Ireland, a 3.0 MW wind
turbine completed in early 2016 is the latest project in a series of energy use and
CO2 emission reduction efforts ongoing since 2006, which through the end of
2015 have reduced the energy cost and CO2 emissions at that location by nearly
30 percent. (Johnson & Johnson, 2016)
In these two examples, the vague references a series of and a range of do not
over-elaborate on the exact figures accounting for the company’s CSR
performance. Internal and external audiences are presented with fuzzy
accounts, with only an appropriate amount of information disseminated,
indicating the company’s operational and strategic decisions.
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 87
Table 2. Vague items associated with “quantity”.
6.2. Vague items associated with “degree”
The second group of vague expressions is associated with degree such as
important, well, significant, better, appropriate, effective, and good, which follows
“quantity” in terms of frequency. The high occurrence of these positive
evaluative items echoes previous research showing that companies seek to
gain trust from stakeholders (Fuoli, 2017). The company attempts to
exemplify organisational competence and rational justifications for its
decisions or prospects through adjectives and adverbs such as important, well,
and significant, and it demonstrates readiness to carefully assess impacts of
planning and future scenarios through adjectives such as relevant and
appropriate.
BIxI JIn
Ibérica 43 (2022): 77-10288
BIXI JIN
Ibérica 43(2022): …-…
Lexical item Freq Freq per thousand
more than 760 1.09
many 434 0.62
various 392 0.56
some 294 0.42
several 210 0.30
approximately 205 0.29
nearly 139 0.20
amount of 137 0.20
a variety of 122 0.17
millions of 107 0.15
at least 103 0.15
a range of 84 0.12
numerous 78 0.11
much 62 0.09
almost 62 0.09
a series of 55 0.08
thousands of 53 0.08
or more 38 0.05
most of 33 0.05
a number of 30 0.04
Table 2. Vague items associated with “quantity”.
6.2. Vague items associated with “degree”
The second group of vague expressions is associated with degree such as
important, well, significant, better, appropriate, effective, and good, which
follows “quantity” in terms of frequency. The high occurrence of these positive
evaluative items echoes previous research showing that companies seek to gain
trust from stakeholders (Fuoli, 2017). The company attempts to exemplify
organisational competence and rational justifications for its decisions or prospects
through adjectives and adverbs such as important, well, and significant, and it
demonstrates readiness to carefully assess impacts of planning and future
scenarios through adjectives such as relevant and appropriate.
Example 8. Through this work, Colgate will be an important catalyst to increase
the transparency of supplier companies in the oleochemical industry.
(COLGATE, 2017)
Example 9. Although we missed our Healthy Future 2015 goal to achieve a 10
percent absolute reduction in water use by 2015, with our 2010 consumption as
a baseline, compared to our performance in 2013 and 2014 we made s ig n if ic a nt
advancement in water reduction in 2015. (Johnson & Johnson, 2015)
Example 10. L’Oréal convened a panel of 12 international experts, who
contributed to developing the methodology – in line with European Product
Environmental Footprint standards – in order to ensure that the final methodology
would lead to a relevant and robust assessment of impacts. (L’Oréal, 2018)
In Example 8, the drafter used an important catalyst to describe Colgate’s role
in transparency improvement. however, the use of important is vague here
because there are no specific criteria to delimit whether corporate efforts are
important. Likewise, the use of significant in Example 9 also makes the
meaning of the sentence fuzzy. Significant is used to describe the large
number of products shipped by the company, demonstrating its overall
power in production.
Example 10 uses very and well to cover a wide range of staff health check-up
issues informed by. In the absence of facts and data, very and well can be seen
as a strategic use to praise their social responsibility behaviours, aiming to
construct a positive and responsible image.
Moreover, in Table 2, it is apparent that most of these vague expressions
(e.g., important, well, significant, good, appropriate, effective, and great) are degree
markers. The companies made more discursive efforts to demonstrate the
current situation of socially responsible behaviours by addressing
stakeholders’ concerns and needs through concrete language expressions.
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 89
Table 3. Vague items associated with “degree”.
6.3. Vague items associated with “time”
Another major group of vague expressions is associated with “time” or
“frequency,” as demonstrated by the use of expressions in this group,
including early, recent, often, by the end of, and always. Compared to the other
three categories, it seemed that they were rarely observed. This might be
explained by the fact that CSR reports are often overtly promotional with the
frequent use of positive evaluative expressions (Livesey & Kearins, 2002).
Vague time references have been found to be prevalent in legal texts for
other reasons (Li, 2017). here, they appear to play a specific part related to
management of the corporate image. given CSR’s role in both reporting the
regular information over the year and predicting the future, they provide
leeway for the company to deal with strategic decision making or fulfil its
commitment.
BIxI JIn
Ibérica 43 (2022): 77-10290
BIXI JIN
Ibérica 43(2022): …-…
Lexical item Freq Freq per thousand
important 377 0.54
well 356 0.51
significant 328 0.47
better 283 0.40
relevant 217 0.31
appropriate 173 0.25
effective 161 0.23
very 158 0.23
main 134 0.19
good 106 0.15
necessary 102 0.15
closely 95 0.14
essential 92 0.13
general 92 0.13
significantly 88 0.13
great 84 0.12
vital 83 0.12
fundamental 72 0.10
Table 3. Vague items associated with “degree”.
6.3. Vague items associated with “time”
Another major group of vague expressions is associated with “time” or
“frequency,” as demonstrated by the use of expressions in this group, including
early, recent, often, by the end of, and always. Compared to the other three
categories, it seemed that they were rarely observed. This might be explained by
the fact that CSR reports are often overtly promotional with the frequent use of
positive evaluative expressions (Livesey & Kearins, 2002). Vague time references
have been found to be prevalent in legal texts for other reasons (Li, 2017). Here,
they appear to play a specific part related to maintaining the corporate image:
given CSR’s role in both reporting the regular information over the year and
predicting the future, they provide leeway for the company to deal with strategic
decision making or fulfil its commitment.
Example 11. Efforts in 2015 include two fuel cell projects in California, and
the construction of a new windmill project in Ireland, which was completed in early
2016. (PROCTER & GAMBLE, 2015)
Example 12. In Japan’s cosmetic market, which is now mature, new entrants
have entered from other industries in recent years, and, as a result, competition has
intensified. (KOSÉ, 2018)
Example 13. Many employees cycle to work, which is good for the environment
and their fitness; however, they often ride without helmets, although these are vital
for their personal safety in rush-hour traffic. (Beiersdorf, 2016)
In Example 11, the vague item early is related to the time range. The start and
end dates of ea r l y 2 0 1 6 are not clarified. E ar l y 2016 in a general sense may
refer to the first two or three months of 2016. It is unnecessary to specify
the exact date of completion of the new windmill project because e ar l y 2016
as a time reference would not prohibit the smooth understanding of target
readers. Besides, it is very likely that the drafter would not know the exact
date.
Similarly, Example 12 demonstrates the vague term recent, which can refer to
several years ago or last year. A possible explanation for the use of recent is
that it is difficult to determine which year new entrants from other industries
enter Japan’s market, which is a continuous process.
In Example 13, often suggests that the frequency of employees who ride
bicycles without helmets is relatively high, even though the specific rate of
those who ride without helmets may not be officially counted by the
corporation. In the absence of specific data and information, it is wise to use
vague items like often to express the same meaning. Moreover, the company
uses often to emphasise the seriousness of current safety problems,
highlighting the role of the safe cycling project launched by the company.
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 91
Table 4. Vague items associated with “time”.
6.4. Vague items associated with “softening stance-taking”
VL in this category is manifested by the use of perspective-taking verbs
(suggest, think) and possibility indicators such as possible and may. The
emergence of these vague items may reflect a general tendency that CSR
reports are subjective. These linguistic manifestations project reasoned
justification for company decision and engagement in CSR activities, while
also leaving adequate space for negotiations in the case of unforeseen
circumstances or negative evaluations, thus revealing the “dialogic nature” of
corporate discourse (Fuoli, 2017). The examples below may be interpreted
as a responsive discursive strategy aiming at displaying their careful
engagement with social and environmental performances.
Example 14. Although the quality of Kosé’s products is world-class, I think it
is crucial to carry out branding that achieves universal global acceptance. Kosé will
do this not by giving up its unique qualities but by enhancing them as it flexibly
responds to changing social conditions. (Kosé,2019)
Example 15. In 2018, OHSAS 18001 will probably be replaced by the
international standard ISO 45001, and we already integrated the expanded
requirements associated with this into our internal checklist for our ESMAS
audit program in 2016… (LVHM, 2017)
In Example 14, compared to the likelihood verb believe expressing a strong
sense of obligation and accountability, think conveys a weaker degree of
BIxI JIn
Ibérica 43 (2022): 77-10292
BIXI JIN
Ibérica 43(2022): …-…
Lexical item Freq Freq per thousand
early 155 0.22
recent 100 0.14
often 98 0.14
by the end of 89 0.13
always 87 0.12
recently 47 0.07
at the end of 30 0.04
usually 16 0.02
sometimes 16 0.02
frequently 11 0.02
Table 4. Vague items associated with “time”.
6.4. Vague items associated with “softening stance-taking”
VL in this category is manifested by the use of perspective-taking verbs (suggest,
think) and possibility indicators such as possible, may. The emergence of these
vague items may reflect a general tendency that CSR reports are subjective. These
linguistic manifestations project reasoned justification for company decision and
engagement in CSR activities, while also leaving adequate space for negotiations
in the case of unforeseen circumstances or negative evaluations, thus revealing the
“dialogic nature” of corporate discourse (Fuoli, 2017). The examples below may
be interpreted as a responsive discursive strategy aiming at displaying their careful
engagement with social and environmental performances.
Example 14. Although the quality of Kosé’s products is world-class, I think it is crucial to
carry out branding that achieves universal global acceptance. Kosé will do this not by
giving up its unique qualities but by enhancing them as it flexibly responds to changing
social conditions. (Kosé,2019)
Example 15. In 2018, OHSAS 18001 will probably be replaced by the international
standard ISO 45001, and we already integrated the expanded requirements associated with
this into our internal checklist for our ESMAS audit program in 2016… (LVHM, 2017)
In Example 14, compared to the likelihood verb believe expressing a strong sense
of obligation and accountability, think conveys a weaker degree of engagement.
Thus, the statement signals an emerging ambition on the part of Kosé to promote
its branding globally, but also leaves the possibility open if Kosé fails to fulfil this
expectation.
In Example 15, the use of probably implies the drafter’s unwillingness to commit
him/herself concerning whether the statement is true or false. However, VL here
engagement. Thus, the statement signals an emerging ambition on the part
of Kosé to promote its branding globally, but also leaves the possibility open
if Kosé fails to fulfil this expectation.
In Example 15, the use of probably implies the drafter’s unwillingness to
commit him/herself concerning whether the statement is true or false.
however, VL here gives stakeholders a clear message that companies are
actively adjusting themselves to cope with changes given the relevant
standards. Meanwhile, LVhM leaves some room for itself in case what they
predict does not happen.
Table 5. Vague items associated with “softening stance-taking”.
6.5. Communicative functions of vague language
From the above discussion, it can be seen that the drafters of CSR reports
can benefit from the use of VL. overall, five major situations can be
summarised as to how drafters tend to strategically use VL, with adaptations
based on Li (2017) and Zhang’s (2018) work: to give an appropriate amount
of information, demonstrate solidarity, enhance persuasion, and self-
protection.
a) To give an appropriate amount of information. In drafting a CSR text,
writers are expected to provide relevant information on CSR-related
development as an annual highlight to stakeholders for future
investment, on the one hand. on the other hand, there will always be
unexpected circumstances or future contingencies so that companies
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 93
A CORPUS-ASSISTED STUDY OF VAGUE LANGUAGE IN CSR REPORTS
Ibérica 43 (2022): …-…
gives stakeholders a clear message that companies are actively adjusting
themselves to cope with changes given the relevant standards. Meanwhile, LVHM
leaves some room for itself in case what they predict does not happen.
Lexical item Freq Freq per thousand
probably 228 0.33
possible 201 0.29
may 134 0.19
assume 109 0.16
estimate 94 0.13
think 38 0.05
might 29 0.04
suggest 21 0.03
Table 5. Vague items associated with “softening stance-taking”.
6.5. Communicative functions of vague language
From the above discussion, it can be seen that the drafters of CSR reports can
benefit from the use of VL. Overall, five major situations can be summarised as to
how drafters tend to strategically use VL, with adaptations based on Li (2017) and
Zhang’s (2018) work: to give an appropriate amount of information, demonstrate
solidarity, enhance persuasion, and self-protection.
a) To give an appropriate amount of information. In drafting a CSR text, writers
are expected to provide relevant information on CSR-related development as an
annual highlight to stakeholders for future investment, on the one hand. On the
other hand, there will always be unexpected circumstances or future contingencies
so that companies are unable to inject precise details and thus resort to VL for self-
protection. There may even be a genuine lack of precise information on the part of
corporate management. The vague expressions some or many in Examples 2 and
4, on the one hand, suggest that companies pay attention to protecting the
information that is crucial to cover themselves. While precise information can be
obtained if corporations spend money and time collecting missing information, the
information-seeking process may result in an imbalance between investment and
economic return. It is thus likely that unclear referencing is an effective cost-
saving strategy. A similar finding can be seen in Li’s (2017) study on VL in legal
texts, in which she argued that draftsmen may not have time to check and access
exact information, which thus results in the emergence of VL.
b) To enhance persuasion. In this corpus, we can see that drafters often use vague
items associated with stance markers (e.g., important, very, essential). VL in this
are unable to inject precise details and thus resort to VL for self-
protection. There may even be a genuine lack of precise information
on the part of corporate management. The vague expressions some or
many in Examples 2 and 4, on the one hand, suggest that companies
pay attention to protecting the information that is crucial to cover
themselves. While precise information can be obtained if
corporations spend money and time collecting missing information,
the information-seeking process may result in an imbalance between
investment and economic return. It is thus likely that unclear
referencing is an effective cost-saving strategy. A similar finding can
be seen in Li’s (2017) study on VL in legal texts, in which she argued
that draftsmen may not have time to check and access exact
information, which thus results in the emergence of VL.
b) To enhance persuasion. In this corpus, we can see that drafters often
use vague items associated with stance markers (e.g., important, very,
essential). VL in this category comprises positive evaluative language
items mainly in the form of adjectives to discursively build up a
committed image, establish a positive relationship with stakeholders,
and underscore their annual contribution to social and environmental
issues (Bondi, 2016; Fuoli, 2017). For instance, Colgate seeks to
demonstrate the active role they lead by signalling a confident attitude,
for example by using important as a stance presentation to persuade
stakeholders (Example 8).
c) To protect one’s image. To build a favorable corporate image,
companies may downplay unfavourable information or soften their
position through strategic use of VL. As shown in Examples 14 and
15, concerning prospects and technological advancement, companies
use VL to soften their stances regarding the blueprint of the product.
Rather than elaborate on the advancement of production standards,
the company presents itself as a rational decision-maker with cautious
attitudes, leaving negotiable space for not fulfilling the objectives.
Also, VL cushions the negative effects of potential criticisms. For
instance, as evidenced by the vague term some in “experienced some
difficulties from excessive store openings due to business expansion”,
some can be construed as a face-saving discursive strategy to maintain
a positive corporate image
BIxI JIn
Ibérica 43 (2022): 77-10294
7. Summary and conclusions
The sections above address the following research questions: What are the
prototypical types of VL in CSR reports of the cosmetics industry? Is the
use of VL in CSR reports related to their communicative purposes? If so,
how? Four types and four communicative functions were illustrated. The
linguistic representations of VL are associated with the semantic categories
“quantity,” “degree,” “time,” and “softening stance-taking.” The
communicative functions underlying the linguistic findings are 1) give an
appropriate amount of information; 2) Enhance persuasion; and 3) Self-
protection. I arrived at these findings through a CAdS approach by
analysing CSR through corpus methods; at the same time, the contextual
analysis surrounding VL use in their contexts sensitised me to the salience of
some of the communicative strategies. The approach helps to uncover how
cosmetics companies communicate socially responsible initiatives through
the discursive constructions of VL in CSRs. The semantic types and
pragmatic functions of VL align with the findings proposed in prior research
(Li, 2017; Zhang, 2018) and confirm some observations made in previous
research on the promotional and subjective nature of CSR reports, as
evidenced by the positive evaluative expressions and softening stance
expressions.
The study also contributes to an enhanced understanding of VL on the part
of business practitioners and business communication scholars, enabling
them to critically understand and assess the VL features in corporate
discourse. Such findings of these categorisations and pragmatic functions
can be transferred into actual drafting practices, such as when a company
drafting CSR reports seeks effective and image-enhancing methods. The
strategic employment of these linguistic resources may manage readers’
impressions and navigate them toward a favorable direction. Beyond the
textual level, given the strategic nature of VL use, the study may bridge the
gap between applied linguists and business practitioners, prompting them to
be aware of the value and feasibility of collaboration.
Some avenues for future research on VL in corporate discourse can be
pursued. notably, in the current study, VL was identified based on prior
studies and sample readings. Several other linguistic manifestations of VL
are not considered in this paper (e.g., conditional sentences, passive
sentences, and epistemic modality). The paucity of these lexico-grammatical
features provides a potential source of inspiration for applied linguists to
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 95
uncover more semantic types and pragmatic functions of VL. Further, while
I focused on the investigation of VL mainly through frequency as conducted
in prior research (Li, 2017; Malyuga & McCarthy, 2018), future research
could focus on a more sophisticated quantitative approach (e.g., collocational
patterns), which expands the scope of the data analysis. In addition, only
English versions of CSRs are included in this research. It is hoped that
future research would focus on the VL in both English and non-English
CSRs and then identify the potential social, cultural, and organisational
factors that may account for the linguistic variation between these two types
of reports.
Article history:
Received 03 October 2020
Received in revised 15 February 2022
Accepted 25 March 2022
References
BIxI JIn
Ibérica 43 (2022): 77-10296
Anesa, P. (2007). Vagueness and precision in
contracts: A close relationship. Linguistica e
Filologia, 24, 7-38.
Arinas, I. (2012). How vague can your patent be?
Vagueness strategies in U.S. patents. Hermes –
Journal of Language and Communication in
Business, 25(48), 55-74.
Baker, P., & McEnery, T. (2015). Introduction. In P.
Baker & T. McEnery (Eds.), Corpora and discourse
studies: Integrating discourse and corpora (pp.
244-265). Palgrave Macmillan.
Biber, D., Johansson, S., Leech, G., Conrad, S., &
Finegan, E. (1999). Longman grammar of spoken
and written English. Longman.
Bondi, M. (2016). “The future in reports:
Prediction, commitment and legitimization in
CSR”. Pragmatics and Society, 7, 57-81.
Bondi, M., & Yu, D. (2017). The generic structure
of CSR reports in Italian, Chinese, and English: A
corpus-based analysis. IEEE Transactions on
Professional Communication, 99, 1-19.
Bhatia, V. K. (2004). Worlds of written discourse: A
genre-based view. Continuum.
Bhatia, B. (2012). The corporate social
responsibility report: the hybridization of a
“confused” genre (2007-2011). IEEE Transactions
on Professional Communication, 55, 221-238.
Channell, J. (1994). Vague language. Oxford
University Press.
Cheng W., & O’Keeffe, A. (2015). Vagueness. In K.
Aijmer & C. Rühlemann (Eds.), Corpus
pragmatics: A handbook (pp. 360-378). Cambridge
University Press.
Choi, A., & Triantis, G. (2010). Strategic
vagueness in contract design: The case of
corporate acquisitions. The Yale Law Journal, 119,
848-924.
Cotterill, J. (2007). ‘I Think He Was Kind of
Shouting or Something’: Uses and abuses of
vagueness in the British courtroom. In J. Cutting
(Ed.), Vague language explored (pp. 97-114).
Palgrave Macmillan.
Cutting, J. (2007). Vague Language Explored.
London: Palgrave Macmillan.
Connor, U., & Upton, T.A. (2004). Discourse in the
professions: Perspectives from corpus linguistics.
John Benjamins.
Ditlevsen, M. G. (2012). Revealing corporate
identities in annual reports. Corporate
Communications: An International Journal, 17(3),
379-403.
Ditlevsen, M. G., Nielsen, A. E., & Thomsen, C.
(2013). Corporate reporting: An integrated
approach to legitimacy claims. Journal of Modern
Accounting and Auditing, 9(12), 1637-1643.
Ecovia Intelligence. (2010). CSR & sustainability:
How the beauty industry is cleaning up.
Evison, J., McCarthy, M., & O’Keeffe, A. (2007).
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 97
‘Looking out for love and all the rest of it’: vague
category markers as shared social space. In J.
Cutting (Ed.), Vague language explored (pp. 138-
157). Palgrave Macmillan.
European Commission (2011). Communication
from the Commission to the European Parliament,
the Council, the European Economic and Social
Committee and the Committee of the Regions: A
renewed EU strategy 2011-14 for corporate social
responsibility (COM (2011) 681).
Flowerdew, L. (2005). An integration of corpus-
based and genre-based approaches to text
analysis in EAP/ESP: countering criticisms against
corpus-based methodologies. English for Specific
Purposes, 24, 321-332.
Fordham, A. E., & Robinson, G. M. (2018).
Mapping meanings of corporate social
responsibility – an Australian case study.
International Journal of Corporate Social
Responsibility, 3(1), 1-20.
Fuoli, M. (2012). Assessing social responsibility: A
quantitative analysis of appraisal in BP’s and
IKEA’s social reports. Discourse &
Communication, 6(1), 55-81.
Fuoli, M. (2017). Building a trustworthy corporate
identity: A corpus-based analysis of stance in
annual and corporate social responsibility reports.
Applied Linguistics, 39(6), 846-885.
Fuoli, M., & Hart, C. (2018). Trust-building
strategies in corporate discourse: An experimental
study. Discourse & Society, 29(1), 514-552.
Gruber, H. (1993). Political language and textual
vagueness. Pragmatics, 3, 1-28.
Handford, M. (2010). The language of business
meetings. Cambridge University Press.
Kushal, S. J. (2011). Business communication.
V.K. Global.
L’Oreal. (2018). Annual report.
Li, S. (2017). A corpus-based study of vague
language in legislative texts: Strategic use of
vague terms. English for Specific Purposes, 45,
98-109.
Liu, Y. (2015). A paradigm for business
communication across cultures: Theoretical
highlights for practice. Tsinghua University Press.
Lin, Y. (2021). Legitimation strategies in corporate
discourse: A comparison of UK and Chinese
corporate social responsibility reports. Journal of
Pragmatics, 177(2), 157-169.
Livesey, S., & Kearins, K. (2002). Transparent and
caring corporations? A study of sustainability
reports by the body shop and royal Dutch/Shell.
Organization and Environment, 15, 233-258.
Maignan, I., Ferrell, O. C., & Ferrell, L. (2005). A
stakeholder model for implementing social
responsibility in marketing. European Journal of
Marketing, 39(9/10), 956-977.
Malavasi, D. (2011). ‘Doing Well by Doing Good’:
A Comparative Analysis of Nokia’s and Ericsson’s
Corporate Social Responsibility Reports. In G.
Garzone & M. Gotti (Eds.), Discourse,
communication and the enterprise: Genres and
trends (pp. 193-212). Peter Lang.
Malyuga, E., & McCarthy, M. (2018). English and
Russian vague category markers in business
discourse: linguistic identity aspects. Journal of
Pragmatics, 135, 39-52.
McCarthy, M. (2019). Vague language in business
and academic contexts. Language Teaching,
53(2), 203-214.
Mellinkoff, D. (1963). The language of the law.
Little Brown & Co.
Partington, A., Duguid, A., & Taylor, C. (2013).
Patterns and meanings in discourse: Theory and
practice in corpus-assisted discourse studies
(CADS). John Benjamins.
Pinkal, M. (1995). Logic and lexicon. Kluwer.
Russell, B. (1997). Vagueness: A reader. In R.
Keefe & P. Smith (Eds.), Vagueness: A reader (pp.
61-68). MIT Press. Australasian Journal of
Philosophy and Psychology, 1, 84-92. (Original
work published 1923).
Ruzaitė, J. (2007). Vague language in educational
settings: Quantifiers and approximators in British
and American English. Peter Lang.
Scotto di Carlo, G. (2015). ‘Weasel Words’ in legal
and diplomatic discourse: Vague nouns and
phrases in UN resolutions relating to the Second
Gulf War. International Journal for the Semiotics of
Law, 28, 559-576.
Skulstad, A. (2008). Creating a ‘green’ image in
the public sphere: Corporate environmental
reports in a genre perspective. In R. Wodak & V.
Koller (Eds.). Handbook of communication in the
public sphere (pp. 181-201). De Gruyter.
Swales, J. M. (1990). Genre analysis: English in
academic and research settings. Language.
Cambridge University Press.
van Riel, C. B. M., & C. Fombrun (2007).
Bixi Jin received her doctoral degree in English language Education at the
university of hong Kong. She is now affiliated with the School of Foreign
Languages, Beihang university. her research interests include corpus
linguistics, discourse analysis, and computational social science. her recent
work appeared in IEEE Transactions on Professional Communication and Ibérica.
NoteS
1 The relationship between vagueness and ambiguity is far more complex and the distinction between the
two linguistic phenomena may be blurred, but this is beyond the scope of the current study.
2 Or other, or more, and or any were identified as the VL clusters through n-grams in Sketch Engine.
3 The frequency of each vague item (normalised to per thousand word) obtained from the current study
may not be directly comparable to the frequencies in previous studies (normalised to per million word).
BIxI JIn
Ibérica 43 (2022): 77-10298
Essentials of corporate communication:
Implementing practices for effective reputation
management. Routledge.
Widdowson, H. G. (2004). Text, context, pretext.
Critical issues in Critical Discourse Analysis.
Blackwell.
Zhang, G. (2011). Elasticity of vague language.
Intercultural Pragmatics, 8(4), 571-599.
Zhang, G. (2018). Vague language, elasticity theory
and the use of “some”: A comparative study of L1
and L2 speakers in educational settings: linguistic
analysis of some. Bloomsbury Publishing Plc.
Appendix
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 99
A CORPUS-ASSISTED STUDY OF VAGUE LANGUAGE IN CSR REPORTS
Ibérica 43 (2022): …-…
Appendix
Rank Word Freq
1 the 33476
2 and 31387
3 of 23892
4 to 21888
5 in 18958
6 our 13426
7 a 10643
8 for 9223
9 we 8087
10 with 7434
11 on 5799
12 is 5256
13 by 4816
14 as 4758
15 are 4449
16 button 4421
17 that 4024
18 this 3285
19 sustainability 3280
20 from 3199
21 at 3125
22 health 2855
23 employees 2744
23 products 2650
25 have 2489
26 an 2437
27 management 2430
BIXI JIN
Ibérica 43(2022): …-…
28 their 2383
29 business 2373
30 all 2310
31 more 2234
32 its 2156
Table A1. Top 50 words from the wordlist of the CSR (cosmeics industry).
Table A1. Top 50 words from the wordlist of the CSR (cosmeics industry).
BIxI JIn
Ibérica 43 (2022): 77-102100
BIXI JIN
Ibérica 43(2022): …-…
32 its 2156
33 water 2145
34 global 2102
35 people 2049
36 Johnson 2029
37 or 1936
38 also 1920
39 new 1913
40 has 1906
41 group 1809
42 through 1748
43 environmental 1733
44 safety 1659
45 product 1642
46 which 1619
47 Colgate 1580
48 social 1567
49 care 1563
50 program 1541
Table A1. Top 50 words from the wordlist of the CSR (cosmeics industry).
A CoRPuS-ASSISTEd STudy oF VAguE LAnguAgE In CoRPoRATE RESPonSIBILITy REPoRTS oF ThE CoSMETICS InduSTRy
Ibérica 43 (2022): 77-102 101
A CORPUS-ASSISTED STUDY OF VAGUE LANGUAGE IN CSR REPORTS
Ibérica 43 (2022): …-…
Rank Word Keyness
1 Colgate 1258.82
2 Henkel 1072.857
3 GRI 709.54
4 Kosé 694.74
5 Amorepacific 694.304
6 LVMH 591.087
7 Beiersdorf 490.015
8 tesa 303.321
9 Gorsky 207.69
10 sustainability 148.45
11 Maisons 134.657
12 CDP 125.034
13 RSPO 120.58
14 Innisfree 117.819
15 EHS 116.228
16 UNGC 110.451
17 Janssen 98.977
18 Sekkisei 98.587
19 Lauder 97.139
20 KRW 96.398
21 GHG 95.904
22 Estée 93.972
23 Dior 85.806
23 Hennessy 79.975
25 Decorté 79.943
26 handwashing 78.545
27 supplier 76.195
28 earthwards 74.601
29 button 73.426
BIXI JIN
Ibérica 43(2022): …-…
30 BSBF 72.252
31 recyclability 71.146
32 stewardship 70.529
33 parfums 70.351
34 CSR 69.849
35 Sephora 69.783
Table A2. Top 50 words from the keyword list of the CSR (using the English Web 2020 in Sketch Engine as the
reference corpus).
Table A2. Top 50 words from the keyword list of the CSR (using the English Web 2020 in Sketch Engine as the
reference corpus).
BIxI JIn
Ibérica 43 (2022): 77-102102
BIXI JIN
Ibérica 43(2022): …-…
34 CSR 69.849
35 Sephora 69.783
36 G4-DMA 69.035
37 Nivea 68.187
38 Guerlain 66.518
39 FY16 64.991
40 Credo 64.936
41 Berluti 61.269
42 foreword 61.239
43 ELC 60.923
44 Co2e 60.141
45 packaging 60
46 sustainably 56.277
47 Ecovadis 55.83
48 Terracycle 53.645
49 deforestation 53.129
50 Johnson 52.889
Table A2. Top 50 words from the keyword list of the CSR (using the English Web 2020 in Sketch Engine as the
reference corpus).