Evaluative Language in English Job Advertisements in Diachronic Perspective

The received view has it that genres are subject to historical changes with respect to their functional and language features. The purpose of this article is to examine changes in the evaluation language employed in the genre of British job advertisement, in this way revealing shifts and developments in this type of discourse practice. Drawing on evidence from the corpus of job advertisements published in the Times, the national British newspaper, in the period between 1896 and 2006, this paper proposes an analysis of evaluative language usage in three collections of job advertisements from late nineteenth, middle twentieth centuries, and the 2000s. The regularities present in the data were calculated manually for every period under consideration. The comparative analysis of the data obtained for each of the synchronic layers introduces diachronic perspective into the study by revealing changes in evaluative language throughout the periods as they are reflected in an average ratio of adjectives per text, distribution of evaluative and descriptive adjectives, frequency of the entities evaluated and semantic variation of the adjectives used to evaluate entities. The aim of the study is to understand the changing role of evaluation in the genre, the way evaluative adjectives work, the role they play in persuading the applicant.


Background
The genre of the job advertisement occurs every time when a company or organization feels the need to recruit suitable employees that meet their requirements to fulfill the company's demands.However, research into linguistic features of job advertisements still remains rather scarce, both in synchronic and diachronic perspectives (Rafaeli & Oliver, 1998;Gillaerts, 2012).Linguists mainly concentrate on other members of the set of business genres, such as resumés and letters of application (DeKay, 2003;Yartsev, 2012), leaving job advertisements out of focus of attention.According to the data available, diachronic approach to the study of job advertisements was applied in the studies by Gillaerts (2012), Walters and Fage-Butler (2014).Describing the evolution of this genre in the Flemish newspaper De Standard over a period of sixty years , Gillaerts (2012) focuses on the analysis of its constituent moves, which he discovers remain stable over the period, and reveals the growing significance of evaluation and interaction in this genre, which, as he proposes, indicates the genre marketization.Diachronic research conducted by Walters and Fage-Butler (2014) concentrates mainly on the developments within the job advertisement format in Danish newspapers .The authors conduct the analysis with a clear practical purpose to suggest a format for current job advertisements that have a potential application value for human resource managers.In the present paper we adopt a diachronic perspective to the genre of job advertisement in the British national newspaper the Times covering a period over a century  with the purpose of addressing changes in evaluative language employed with emphasis on frequency of entities mentioned in the texts that are subject to evaluation and semantics of adjectives used for evaluation.

Framework of the Study
According to Thompson and Hunston (2000) evaluation is the broad cover term for the expression of the speaker or writer's attitude or stance towards, viewpoint on, or feelings about the entities or propositions that he or she is talking about.Evaluation is considered a gradual phenomenon, and it may be present to various extents in different text types and genres.Based on the category of evaluation Philip Shaw (2006) suggests that discourse genres could be divided into interested and disinterested ones.According to this division, an interested genre suggests a personal, sometimes manipulative, perspective of the situation rather than an objective and balanced view provided by "disinterested" genres, the authors of which "are expected to be impartial, basing their conclusions or recommendations on all the facts as known to them" (Shaw, 2006).Thus, the difference between the two categories lies in the usage of means of evaluation: Texts belonging to interested genres are more prone to accept evaluative lexis, the evaluation is mostly extremely positive.Evaluation is viewed as an inalienable feature of many genres, for example, advertisements, promotional genres, personal diaries, reviews all are normally evaluative, while some genres virtually exclude evaluation, namely scientific articles, documentation are not expected to translate emotion, subjectivity or evaluation, as they are supposed to be the direct opposite of subjectivity.We assume that besides variation across genres, the evaluative language may also vary within one genre in the course of its historical development thus it might indicate changes in genre conventions.
Although evaluation can be expressed by different parts of speech throughout this paper we have applied the term "evaluation" to adjectives that are traditionally considered as explicit means of expressing evaluation.
Being the largest open word class in English after nouns and verbs (Leech, 1989), adjectives have a huge potential to convey emotion and meaning.In literature adjectives are classified according to various criteria (Marzá, 2012).A number of semantic, syntactic and morphological criteria are used in classifications proposed by various authors, for instance, central and peripheral (Quirk et al., 1972), or emotional and rational evaluation (Starostina, 2013).Numerous semantic classifications of adjectives are proposed in literature that are based on concrete or abstract semantic criteria with the number of semantic classes varying from ten (Dixon, 1982) to twenty four (Lee, 1994).
For the purposes of our study, we adopted a reduced categorization that divides adjectives into evaluative and descriptive, the former conveying the speaker's attitude to the referent (the perceived qualities), the latter stating the "objectively present" (factual or absolute) qualities.For example, "ambitious" is considered an evaluative adjective since it is intended to express the idea that not only is the referent determined to succeed but also that the user is likely to think this is a good thing, while the adjective "furnished" is descriptive, for no obvious connotation is observed.Evaluative adjectives, in turn, can be divided into adjectives of general evaluation (generally-positive and generally-negative) and adjectives of specific evaluation (positive or negative), as proposed by Arutunova (1988).The adjectives that are generally-evaluative characterize the referent indefinitely, by qualifying it as generally good or bad, to various degrees ("excellent", "terrible", "good", "bad", etc. are examples thereof).The other type of evaluation, adjectives of specific evaluation, characterizes a certain facet of the referent rather than it in general ("beautiful", "intelligent", etc.).In turn, specific evaluative adjectives found in job advertisements can be categorized into semantic groups on the semantic basis.For example, the adjective "ambitious" is defined in Merriam-Webster On-line Dictionary as "having a desire to be successful, powerful, or famous", a definition for the adjective "motivated" can be derived from the verb "to motivate" and the noun "motive" defined as "having a need or desire to act".Thus, the common semantic component a "desire to be or to do smth" enables us to include these words into one semantic group.Following Hewing's (2004) terminology, in this paper, evaluative adjectives have been analyzed with respect to evaluated entities in British job advertisements.Drawing on works of different authors Meurs van, Korzilius and den Hollander (2006) enumerate the following textual units that are frequently found in job advertisements: job description, job requirements, the offer (salary, bonus and other incentives), the application procedure (including contact information).According to Rafaeli and Oliver (1998) the elements that are present in most job advertisements are the following: 1) an organizational identity, 2) its human resource needs, 3) information about what is required to fulfill these needs, 4) information about how to contact the organization.As our purpose is to investigate evaluative language we assume that contact information is unlikely to convey any evaluation and we exclude this element from our analysis.On the contrary, as our preliminary observations showed, the company or organization always presents in the text explicitly and may receive evaluation that can be described in terms of self-presentation/self-evaluation by which we understand the construction of a positive image of self for an intended audience.So, we adopt the scheme that includes three entities that are expected to receive evaluation in job advertisement, namely the company/organization, the vacancy, the requirements for a potential employee.We assume that evaluative language of job advertisements is subject to historical changes with respect to the intensity of evaluation, entities evaluated and semantics of evaluative adjectives.

Materials
For the purpose of our study we randomly selected two hundred and ten job advertisements published in the Times, the oldest British national paper on the current market.As we intend to trace diachronic changes in evaluative language in the genre we limited ourselves by sticking to a single newspaper with a relatively stable audience throughout the period investigated so that all language changes could be attributed to the time factor.The job advertisements belong to three years, namely 1896, 1956 and 2006 and distributed equally, seventy of the advertisements were selected for every year from random issues.We intentionally chose periods with a lapse of approximately fifty years so that genre developments, if any, could be traced.

Procedure
To understand the role of evaluation in the genre of job advertisement and whether the genre displays a trend for an increase in evaluation the following procedure was applied.First, we counted an average number of evaluative adjectives per text for each of the synchronic layers in the texts from our corpus and calculating a ratio of adjectives to the total number of words in the text.At the next stage, we counted manually the incidence of evaluative and descriptive adjectives for each period to answer the question whether evaluation or description prevails and how the genre has developed in this aspect.We excluded certain types of adjectives as they are not used in their evaluative or descriptive roles and their admission to the corpus may distort the results.Examples of the adjectives left out of the focus include such adjectives as "general" (in "general accountant"), "managing" (in "managing clerk"), (since they don't describe the noun but rather form a noun phrase), or "previous" (as in "previous situation") since they don't fulfil the qualifying function.Since we have singled out three possible entities that can be evaluated: the adjectives in job advertisements can be used to describe the prospective candidate, mainly the requirements that he/she should meet, the vacant post (to which the advertisement is actually devoted), the addresser (the company/other body that gives the advertisement), we counted the occurrence of entities that receive evaluation for every period examined.Then, we identified adjectives related to the specific entity and compiled lists of the adjectives most frequently used to characterize each of the entity considered for each year.We only included the adjectives that were used twice or more.After that, we categorized the adjectives into semantic classes and traced the changes in the percentage of the adjectives belonging to each semantic class in order to capture the semantic variation in the qualification of the three evaluated entities.
By comparing the data obtained for each of the periods under investigation we identified shifts in distribution and semantics of adjectives and interpreted these shifts in relation to the sociocultural context of each period.

Changes in the Number of Adjectives per Text in British Job Advertisements
The percentage ratio of adjectives (used in relation to all the entities they evaluate) and the general word count in the advertisements is represented in Table 1.Our results show that both the absolute number of adjectives per text (from 0-3 in 1896 to 3-10 in 1956 to 5-6 in 2006) and the percentage of adjectives per text have increased since mid-century.The length of the advertisement has also grown (from 17-27 words in 1896 to 53-130 in 1956 to 94-100 words in 2006).an average number of adjectives per text 7.9 % 7.6 % 11.7 % Thus, an average number of adjectives per text was at 7.9 %, 7.6 %, and 11.7 % in 1896, 1956 and 2006, respectively.The percentage of the evaluative adjectives remained stable from 1896 to 1956, however, there is a remarkable trend in 1956 to substantivize the qualities that would probably have been expressed by adjectives in 1896 (e.g. using the noun "enthusiasm" in 1956 while in 1896 the adjective "enthusiastic" would be expected).Such cases fall outside the realm of our research (since the latter is devoted to evaluation expressed through adjectives), so the limitations of the current research did not allow us to track this phenomenon with higher precision.The incidence of evaluative adjectives rose significantly by the beginning of the 21 st century, with the companies wanting to appeal to the emotions of the applicant and encourage him/her to apply for the job.

Evaluation vs Description in British Job Advertisements in Diachronic Perspective
The study revealed that the incidence of evaluative adjectives in the total number of adjectives has grown (even though it had been initially high) throughout the period in question, the number of evaluative adjectives has increased in general as compared to descriptive adjectives.
Figure 1.The distribution of evaluative and descriptive adjectives related to each referent Figure 1 shows the distribution of evaluative and descriptive adjectives in the job advertisements related to the entities we investigate.The bars represent the percentage of evaluative adjectives (as opposed to descriptive ones) in relation to the total number of adjectives used to characterize the referent.The data obtained show two kinds of changes: the dynamics in the use of evaluative and descriptive adjectives and the total number of adjectives found in the corpus.We would like to point out that we regarded any recurrent use of an adjective as a separate instance of use thereof.
Despite a clear increase in prevalence of evaluative adjectives the leap over the twentieth century is insignificant as the percentage of evaluative adjectives was initially high in the year 1896.The frequency in the number of evaluative adjectives used in the vacancy description is the highest in 1956, as the employers became more inclined to advertise the image of the vacancy rather than the objective information on the vacancy and feel the need to appeal to the emotions of the addressee.A slightly less dramatic growth in the incidence of evaluative adjectives is seen in the description of requirements for candidates in 2006 as compared to 1896 (by 11.04%), which is also related to the recognition of the need to make stronger impression on the candidate in order to increase effectiveness.Overall, the job advertisement genre towards the 2000s has acquired more properties of an actual advertisement rather than of a business document, thus displaying the characteristics of an "interested" genre.This conclusion is in good agreement with the findings obtained by Gillaerts (2012).There are also certain alterations in the way the referents qualified by adjectives are evaluated.As can be seen in Figure 3, there is not a single referent that is characterized by adjectives in every advertisement (there will always be advertisements where one or more referents are not mentioned at all).However, as is demonstrated in Figures 2 and 3, the reference to the addressee became overbearing in 2006 (the candidate was described in 66 out of 70 advertisements; and those 66 contained 107 adjectives, which means that the candidate is normally described using at least two adjectives) and in 1896 (78 adjectives in 43 texts, which also means the density of 1-3 adjectives to characterize the addressee in a smaller amount of texts than in 2006).The number of adjectives characterizing the addressee is lower in 1956 ( 53) and is used more sparsely (roughly one per advertisement).
The reference to the vacancy became more frequent in mid-century both in terms of the number of adjectives used and advertisements employing adjectives to qualify the vacancy.In the 2000s vacancy evaluation dropped to the level comparable to late nineteenth century.As for the reference to the company, it has shown steady growth in adjectival qualification, both in the number of adjectives used in total and in the number of advertisements containing such qualification.So, the pattern of distribution of adjectives in relation to entities (candidate and vacancy) in late nineteenth century and the 2000s bear strong resemblance, which is one of the many examples of similar trends these periods have.

Semantics of Evaluative Adjectives in British Job Advertisements
To understand how semantics of evaluative adjectives have changed, we have compiled lists of adjectives most frequently found in job ads, one list for each year and combined them into tables that show adjectives used for the description of a certain entity in a certain year (3 referents, 3 milestone years), the total number of adjectives is stated in the top portion.

Evaluation of Requirements for Candidates in British Job Advertisements
The description of the requirements for the prospective employee is the whole reason to give the advertisement, so it is by far the most elaborate part of the advertisement with all of the tendencies being most pronounced in it.
The descriptive adjectives used in reference to the addressee usually describe skills (1956,2006: "familiar with", "essential") or social, gender or other characteristics ("married/unmarried", "unconnected with the press", 1896).Table 2 represents the adjectives describing the requirements a candidate for the job should satisfy (Table 2).As seen from the list of adjectives on the following table, all of the frequently used adjectives are evaluative.in 2006, while in the latter period we can observe a tendency towards using a wider selection of adjectives less often (e.g."proactive", "outstanding", "tenacious", each used only twice).Thus, the list of most frequently used evaluative adjectives in earlier years used to be comprised of a limited number of adjectives used extensively (circa 55% of all the cases of usage), the remaining 45% being the adjectives used once.So, in later years the adjectives became more semantically meaningful (a larger percentage of them refers to specific qualities rather than expresses generally positive evaluation) and are chosen from an extended set of typically used adjectives (the number of adjectives used more than once has grown from 10 as of 1896 to 20 in 2006).
The positively evaluated qualities expressed by adjectives have obviously changed throughout the twentieth century (in 1896 "gentlemanliness" and the quality of being "well-versed" were required by some employers, in 1956 the quality of being "sound" was listed as necessary, while in 2006 "competitiveness" and "proactivity" were demanded).This choice of adjectives reflects the values of the time (industrial society being focused on respectability ("gentlemanliness"), while later the accelerated speed of production called for dynamism that is expressed by the adjectives such as "proactive", etc.).
We have singled out the following semantic classes of adjectives describing the requirements for a potential candidate for the vacancy offered and compared the frequencies of semantically categorized adjectives: ambitiousness represented by the adjectives "ambitious", "motivated", "entrepreneurial", "competitive", "proactive", "keen", "articulate", "energetic".We have summed up the incidence of the adjectives of this semantic class for each year: as of 1896 the percentage of the adjectives we have referred to this class is 2.6%, in 1956-7.6%, in 2006-21.7%.
As the list above shows, the potential employee has been characterized from the point of view of ambitiousness at an increasing rate throughout the century (2.6% in 1896 vs. 21.7 in 2006); the addressee's experience/knowledge were desired by the addresser at all times, but the incidence of such adjectives is higher in 1956, after which it goes back to the original rate (11.5% in 1896, 22.7% in 1956, 7.5% in 2006).Diligence and related qualities were least asked of the addressee in the mid-century.And finally, the vague qualifications of the addressee were asked for (probably as a space filler) across the period with no significant fluctuations.

Vacancy Evaluation in British Job Advertisements
Table 3 contains information about the adjectives that are used in the description of the vacant post being advertised.As in the previous case, the most frequently used adjectives are evaluative, the descriptive being used to characterize the working conditions ("furnished"; "quiet") and the period of time over which the vacancy will be closed or the worker will be employed ("immediately"; "short/long-term").Since the qualification of the vacancy is multifaceted (that is, there is a number of properties of the vacant post that can be qualified and/or evaluated, e.g. the prospects, the salary, the overall responsibilities), in this case only we included the nouns to which the adjectives refer in order to give a better idea of the peculiarities of the qualification.As can be seen in the chart, "good" and "liberal" were the adjectives most frequently used in the description of the vacant post in 1896, in the year 1956 the vacancy was typically described as "good", "excellent" and "interesting", while in 2006 the vacancy was mostly referred to as "competitive" and "excellent".
The vacancy is the referent that is least likely to be characterized by adjectives in comparison with the other referents.In contrast to the description of the addressee, the description of the vacancy does not show an increase in the number of adjectives used but does reveal a tendency towards diversification since the percentage of commonly used adjectives has fallen by half (the total of 43% in 1896, peaking in 1956 with 53% and falling in 2006 to 22%).Otherwise, evaluative adjectives used in the description of the vacancy exhibit the same dynamics as the adjectives found in the description of the requirements for the candidate.Namely, the amount of adjectives expressing generally positive evaluation ("good", "excellent") has decreased (in the mid-20 th century advertisements the usage of generally-positive adjectives was the most frequent (37.7%), having plummeted to 6.7% in the 2000s).The range of adjectives used in more recent advertisements show that they somewhat lose the emotional evaluation.The choice of adjectives sometimes depends on what words or concepts are popular in the given period of time: e.g., the qualifications of the vacant post as "unique" or "extensive" in 2006.
Likewise, was common to use the evaluative adjective "large" when describing the working conditions at the end of the nineteenth century, which was non-typical for any other period.The inclination to characterize the vacancy as the one involving responsibility waxed towards the mid-century and waned once again towards the year 2006 (which is opposite to the aforementioned drop in usage of the adjectives of the semantic class "diligent" in reference to the addressee in mid-century).The usage of adjectives that imply the lucrative nature of the post.
The major element that receives evaluation throughout all the periods is salary that is described as "liberal", "competitive", "high", "adequate".The total number of evaluations that this element receives is 10.8% in 1896, 7.2% in 1956, and 6.7% in 2006 thus demonstrating a slight decline by the beginning of the 2000s.

Evaluation of the Company/Organization in British Job Advertisements
At the final stage, we estimated the frequency of adjectives used to evaluate the company/organization that can be regarded as means of self-presentation used by the company for promoting its positive image to persuade a prospective candidate to apply for a job.Table 4 represents the adjectives that are used in the self-presentation of the employer.We identified only two descriptive adjectives that were used frequently enough to be displayed in the chart: the characteristics of size, "large" and "small", while the rest of the list includes only evaluative adjectives.As can be deduced from Table 4, in 1896 "old-established" was the most frequently used adjective, followed by "good" and "important", in 1956 "large" became the most popular evaluative adjective, closely followed by "old-established", and "expanding".The year 2006 retains "expanding" as one of the adjectives most frequently used but "busy", "dynamic" and "progressive" are more typical for self-presentation, besides, the overall amount of adjectives that were used more than once, has doubled.
The following semantic groups have been set out: stability represented by lexical units "well-/long-/old-established", "large".The concept is represented in 25% of all cases in 1896, more than doubled reaching 54.1% in 1956, and 6.2% in 2006; the concept growth: "growing", "expanding" on the contrary is not found in 1896, sharply increases up to 18.2% in 1956, and declines to 8.4% in 2006.The semantic group with the common meaning of impact comprises adjectives "influential", "well-known", "important", "prestigious".The distribution of this group over the periods is the following: 28% in 1896, 9.1% in 1956, and 4.2% in 2006.The last semantic group denoting dynamism expressed in the adjectives "dynamic", "progressive", "busy" does not occur in the previous periods and arises in 27.2% of all adjectives occurrences evaluating the company/organization in 2006.
In keeping with the tendencies outlined above, self-presentation has faced the following changes over the course of the twentieth century: the increase in the total number of adjectives used (from 25 in 1896 to 48 in 2006); decreasing incidence of generally-evaluative adjectives (12% in 1896, but later the descriptions of the addressee are devoid of generally-evaluative adjectives); greater variety of commonly used adjectives (as seen in the list of adjectives evaluating the addressee); the focus of evaluation being dependent on the values and key concepts of the socio-cultural circumstances (the prevalence of the company's self-qualification as "old-(well-)established", in 1896 and 1956 and "dynamic" and "prestigious" in 2006).The latter two facts probably mean that the evaluation is actually poorly adjusted to the specific nature of each individual company and is a conventional gesture.
So, the company's self-presentation as "stable" is common at the end of the nineteenth century, extremely widespread in 1956 and is conspicuously absent in the 2000s.In the meantime, the quality of being "influential" has been consistently losing popularity (from 28% in 1896 down to 4.2% in 2006).The adjectives that share the meaning "dynamic", however, are found exclusively in 2006 in large numbers but no earlier (0% both at the end of the nineteenth century and mid-20 th century).

Conclusion
Let us reiterate and/or examine in greater detail the aforementioned information.Overall, the percentage of evaluative adjectives in relation to the total word count rose by the beginning of the 21 st century.As it was found above, evaluative and descriptive adjectives are not used equally in our corpus, the first being absolutely predominant in the genre in question at all stages of its development.The total number of adjectives used to describe a certain entity has also grown (with the sole exception of the vacancy description wherein fewer adjectives are found), the most noticeable growth being with the requirements for a candidate.This trend is explained with the intention of the recruiters to persuade the potential candidate by appealing to emotions and encouraging him/her to apply for the job.Interestingly, analysis shows that the results collected for 1896 and 2006 in some aspects share more features semantically (but not lexically) with each other than with the year 1956.Namely, the semantic classes of adjectives being used to describe the referents (experience, diligence of the candidate; the responsible nature of the vacancy; the stability and growth of the company).The presentation of the employer (normally the company that offers the vacant post) expressed through evaluative adjectives is also becoming more emotional and appealing which can be explained by the recognized importance of construing the company's image.However, the general trend does not seem to apply to the presentation of the vacancy itself.
Basically, the increased role of evaluation in the description of the entities indicates that the genre possesses more propensities of an advertisement, a purely interested genre, rather than the features of a disinterested genre, even more so towards the 2000s.Figures also demonstrate that generally-positive adjectives like "good" and "excellent" are largely dismissed in the 2000s, probably as a weak option for qualifying the entity properties.However, the evaluation is somewhat semantically vague due to the standardization of sorts, as there is a limited number of adjectives from which to choose when describing the entity.The tendency picks up with the time, and the occurrence of the adjectives from the common list may comprise up to 55% of all adjective usage (in 2006).
Nevertheless, the commonly used adjectives become more evenly spread throughout the corpus.The semantic analysis of the adjectives employed for evaluation reveals values dominating in the periods discussed.In 1896 the priorities are given to experience as a requirement for a candidate, the company's stability and high salary as a benefit to the potential employee while in 1956 the requirement for experience and the company's stability are supplemented by the accent on the company's growth, though the evaluation of the salary remains comparatively high throughout all the periods.In 2006 the focus of evaluation is on the company's dynamism that suggests career prospects and ambitiousness of a prospective candidate and in this way the underlying ideology of the text implies that the interests of the company and the potential employee match.

Future Research
The paper suggests prospects for future research.First of all, we investigated British advertisements only, so the prospective area is to study evaluation patterns in job advertisements placed in other media, mainly electronic papers, social networks, etc., to follow on-going changes since the language of job advertisements is constantly under development.Considering job advertisements written in other cultural environments suggests that comparative studies can be conducted.Furthermore, using a larger data set with British job advertisements from other printed sources of the past periods would shed more light on how evaluation has evolved, as it would help to ensure that the patterns of evaluation found in the present study are consistent and reflect the current trends.Moreover, just as this paper has revealed the time-sensitivity of the genre with respect to evaluation, other pragmatic features of the genre need closer attention in diachronic perspective.

Figure 2 .
Figure 2. The number of texts with evaluative adjectives

Table 1 .
The percentage of evaluative adjectives in relation to the general word count

Table 2 .
The frequency of adjectives evaluating requirements for the candidate in the corpus of job advertisements

Table 2
demonstrates a relative drop in the usage of generally-evaluative (and hence semantically empty) adjectives like "good" and "excellent" from 23.1% of all the instances used in 1896 to 15.21 % in 1956 and 14%

Table 3 .
Frequency of adjectives evaluating the vacancy in the corpus of job advertisements

Table 4 .
Frequency of adjectives evaluating the company/organization in the corpus of job advertisements