Story of Aaker ’ s brand personality scale criticism

Purpose – Aaker’s brand personality scale (BPS) published in 1997 has revived hitherto sluggish interest in brand personality research. With time, the BPS, most cited work in brand personality, also faced criticism across dimensions. This paper aims to review the popular journals published after 1997 for criticism related to BPS. Design/methodology/approach – Papers using Aaker’s BPSwithout change/with change are identified and scrutinized for reasons for the usage of BPS. Papers on brand personality that have avoided BPS are also scrutinized for reasons of avoidance. Independent efforts of understanding brand personality without Aaker’s framework are also reviewed. In-depth study of all these papers is done to report the criticism of Aaker’s BPS. Findings – This review identifies the criticism of BPS and classifies it across six categories – definition, dimension, methodology, concept, words and generalizability related criticism. This paper argues that some issues such as definition, conceptual understanding of brand personality and methodology used to develop BPS need further attention of scholars. On the other hand, issues of dimensions, words used and generalizability can be attributed to evident reasons, such as culture andmeaning given to words because of native language. Originality/value – This criticism and interest in Aaker’s BPS are unprecedented. It has been 20 years since BPS was published. Many scholars have countered the Aaker’s BPS through their work; however, a comprehensive review covering all criticisms and issues of BPS is still missing in literature. This paper is filling this gap in literature.

Palabras clave -Escala de Personalidad de Marca, Jennifer A Aaker, Crítica, Medici on de la Personalidad de Marca, Revisi on de la Personalidad de Marca Tipo de artículo -Trabajo de investigaci on Studies highlighting the relationship between consumers' self and products they use have been receiving attention of researchers since long (Birdwell, 1968;Dolich, 1969).Currently, literature is scrutinizing this relationship by exploring the relationship between brand personality and human personality.Azoulay and Kapferer (2003) argued that this relationship was introduced quite early than scholars believe.The concept of brand personality was coined as early as 1955 (Avis and Aitken, 2015) and has been evolving through contribution from various streams, such as retail, advertising and entertainment.Even the term itself (brand personality) is believed to be coined by advertisers.Since then, the efforts are on, but despite efforts, scholars have failed to generate high-level interest among researchers towards brand personality.The sluggishness in the domain of brand personality was broken by Aaker (1997) when she developed a scale (brand personality scale -BPS) to measure the brand personality.Aaker came out with five dimensions of brand personalitysincerity, competence, excitement, sophistication and ruggedness.Her work has triggered interest among scholars to explore the concept of brand personality.As a result, post 1997, brand personality was tested for its potential relationship with human personality (Aaker et al., 2004), brand associations (Freling and Forbes, 2005), organizations, consumers (Sweeney and Brandon, 2006;Sung and Kim, 2010), preference of consumer (Swaminathan et al., 2009), brand identification (Kuenzel and Halliday, 2010), brand trust, quality (Ha and Janda, 2014), gender (Das, 2015), word of mouth (Tho et al., 2016) and consumer loyalty (Chung et al., 2017).The concept was even extended to other constructs, such as destination personality (Kaplan et al., 2010;Ye, 2012) and corporate brand personality (Keller and Richey, 2006).The popularity of BPS has resulted in blind faith of some scholars to adopt BPS in their studies without modification (Table I).
To understand the advanced scholarly interest about brand personality post 1997, top 20 journals of marketing as per latest available SJR score of 2015 from SCIimago Journal and Country Rank (SJR, 2015) were reviewed for articles published on brand personality in and after 1997.The review resulted in list of 30 articles on brand personality (Table II).All these articles explored the different dimensions of brand personality.Most of these studies cited Aaker (1997).This is enough ground to accept that Aaker's BPS is a seminal work.
The BPS till date is the most cited and applied scale in literature of brand personality.Aaker (1997) can be given credit to popularize the concept but she definitely has not introduced it.Azoulay and Kapferer (2003) called the opening up of pandora box of brand personality researches after Aaker's work as "renewed interest in old concept" (italics in original).Along with acceptance, arguments are also witnessed stating that the work of Aaker requires relook before it matures itself to culmination.As a result, literature has witnessed the contribution from other researchers who have avoided Aaker's BPS and developed new models to measure brand personality.For example, for a niche domain such as sports brand personality, four BPSs (Basket Ball -Carlson et al., 2009;Professional US Sports Club -Braunstein and Ross, 2010;Netball Club -Heere, 2010;Professional Sports Club -Tsiotsou, 2012) are developed (See Table III for the list of studies published on brand personality after 1997, which did not use Aaker's BPS despite its popularity.These studies took reference of other measurement scales to operationalize brand personality.Table III also highlights the reasons for avoiding Aaker's BPS in these studies).These counter BPS efforts did not pick up the limelight as Aaker did, yet scholars have adopted these new models to understand and extend the concept of brand personality to different product categories.
With time, the criticism of BPS has increased multifold.Majority of articles, which are published in recent years, directly or indirectly comment on the applicability of BPS.This criticism and interest is unprecedented.It has been 20 years since BPS was published.Many scholars have countered the Aaker's BPS through their work; however, a comprehensive review covering all kinds of criticism and issues of BPS is still missing in literature.This paper will fill this gap in literature.Hence, this paper reviews the criticism related with Aaker's work as reported in existing literature.All major journals are reviewed for the work published after 1997 relating to brand personality.Work on brand personality prior to 1997 is also reviewed to arrive at a better understanding of brand personality.On the basis of review, criticism relating to Aaker's BPS is divided in six categoriesdefinition related, dimensions related, methodology related, concept related, ethnocentric related and generalizability related.All these issues are detailed in subsequent headings of this paper.

Definition-related criticism
Like any measurement scale, the maturity, acceptance and evaluation of BPS should start from the legitimate understanding of the construct (brand personality) for which the scale is developed.Hence, analyzing the definition used for the construct for any potential issue is a good starting point.Aaker's work is based on her own understanding ofwhat brand personality is? (Definition of construct becomes the base for scale development).Aaker defined brand personality as -"(The set of human characteristics associated with a brand)" (italics in original).This definition is rooted in anthropomorphism (transferring human characteristics to non-humans).It is a legitimate assumption of this definition that all   Aaker's brand personality scale human characteristics can be transferred to brands.Idea of transferring all possible human characteristics to brands has not been accepted by many scholars with assurance.For example, Bishnoi and Kumar (2016) strongly argued that it would be wrong to associate certain traits ("sentimental" and "sincere") of human personality with brands.This applicability changes with the product category.Earlier researchers and scholars also had the same argument (Austin et al., 2003;Azoulay and Kapferer, 2003).Another challenging argument about the anthropomorphism in brand personality and the way it is defined is that the definition is too wide and its scope should be limited to the extent it can be applied to brands (Azoulay and Kapferer, 2003).Subsequently, literature witnessed the introduction of new definition of brand personality -"Brand Personality is the set of human personality traits that are both applicable to and relevant for brands" (Azoulay and Kapferer, 2003) (italics in original).Here, the scope of human personality traits for application on brands is checked and only those traits that can be relevant for brands are considered.Subsequent to these advancements, Ambroise and Valette-florence ( 2010) developed a new scale called brand personality barometer by taking only those traits that are suitable for brands.Sweeney and Brandon (2006) also agreed on the idea adopted by earlier researchers that brand personality relates to human personality traits but they conditioned it to only those traits of human personality that relate to the interpersonal and relationship-based characteristics of human personality and defined brand personality as -"Brand personality is the set of human personality traits that correspond to the interpersonal domain of human personality and are relevant to describing the brand as a relationship partner" (italics in original).Despite these arguments, one thing that is commonly accepted for brand personality is that human personality traits are applicable to brands and hence brand personality as a construct is rooted in the theories of psychology.Though there are exceptions like Heere (2010), who did not agree with the concept and argued that human personality traits are not applicable to brands.But such arguments are miniscule in brand personality literature.The acceptance of the concept of anthropomorphization by researchers, literature and even consumers (Fournier, 1998) weakens the arguments challenging the association of human personality traits with brands.One such example is Sung and Kim (2010), who wrote about the acceptance of anthropomorphization for brands in their paper: [. ..] the notion that inanimate objects such as commercial brands can become associated with human characteristics has been given a considerable amount of attention in marketing and consumer psychology.This symbolic meaning brands acquire is often called brand personality [. ..] (italics in original).
If the popular argument of scholars (Keller, 1998;Eisend and Stockburger-Sauer, 2013) that brands derive their personality through brand encounters and marketers' efforts is true (for example: in India, marketer continuously associates Thums-Up with Salman Khan, who is a youth having attitude and always ready for adventure.This repetitive association makes the consumers to perceive that Thums-Up is having a youthful and adventurous personality.It can be said that the personality traits of human being [Salman Khanyoung, energetic and adventurous] are transferred to non-living being [Thums-Up] and anthropomorphization has taken place.As a result, Thums-Up is said to have acquired a personality), then it is safe to report that anthropomorphization is accepted till the practitioners' level.Scholars like Wee (2004) proved that human personality characteristics can be mirrored for brands on account of anthropomorphization. Bishnoi and Kumar (2016) supported it when they argued that a specific way in which marketers communicated about bike brands is responsible for non-applicability of certain personality traits to bike brands (Indian consumers do not see the bikes as sincere, smooth, independent, successful and sentimental because marketers have never projected bikes so).
Improvisation in definition should be seen as the normal process of advancing the understanding of the construct.For example, the erstwhile definition of "brand" has gone through many changes (Stern, 2006;Jones and Bonevac, 2013) and it can be said that it is still evolving.Brand personality is comparatively a new construct and suggested changes are signs of interest of scholars, which should be taken positively.Pertaining to the above discussion over the appropriateness of definition of brand personality adopted by Aaker, the definition can be called as to be representing the construct of brand personality to a greater extent but still it requires scholars' attention to take the definition towards maturity.Aaker (1997) in her original work came up with 5 dimensions and 15 facets of brand personality.Despite being heavily cited for developing scale to measure brand personality, giving five dimensions and popularizing the multidimensionality of brand personality, Aaker (1997) has been criticized for her dimensions related findings.

Dimensions-related criticism
Aaker is charged of being unable in covering certain dimensions (Alpatova and Dall'Olmo Riley, 2011) of brand personality.Aaker is also charged with fixing the number of dimensions of brand personality.The reason cited for this is heavy dependence of brand personality on personality as construct, which had five dimensions (Davies et al., 2001).Davies et al. (2001) argued that five dimensions of brand personality are not supported by theory.Brand Personality is defined in terms of human personality.Moreover, as discussed earlier, human personality traits are transferred to brands.Taking the traits of human personality to understand brand personality would surely result into some sort of similarity between human personality and brand personality.The charge on Aaker of fixing the number of dimensions may not be substantiated, as Aaker did not prefix the number of dimensions.It is the data on those traits that resulted into five dimensions after appropriate application of statistical tool (factor analysis in this case).
Popularity of Aaker's work has motivated the brand personality community to believe that brand personality has five dimensions.At the same time, there are studies depicting the differences in dimensions of brand personality.Many cross-cultural studies on brand personality depicted the change in dimensions of brand personality.The difference is observed even when Aaker's scale items were used (Table IV).Table IV depicts the studies operationalizing brand personality using traits suggested by Aaker's scale.
Differences are also found when Aaker's items were used with change to measure brand personality.For example: Madrigal and Boush (2008) questioned Aaker (1997) for not including social responsibility in dimensions of brand personality.They argued social responsibility to be an important dimension of brand personality.Anandkumar and George (2011) questioned the applicability of various brand personality dimensions across cultures.In the USA (where Aaker's study was conducted), the emergence of a different dimension (tradition) by Sung (1999) also supported the non-exhaustiveness of Aaker' brand personality dimensions.Studies not only found dimensions to be different from original BPS's dimensions but also questioned the original dimensions proposed by Aaker (1997).Literature has specifically highlighted some dimensions to be reviewed seriously.Out of five dimensions, one that has come under attack very prominently is "ruggedness"."Ruggedness" is termed as the weakest among all dimensions.Scholars have attributed the presence of "ruggedness" in Aaker's list to the decision of taking Levi's Jeans across all brand groups while developing BPS.This is a potential reason for emergence of ruggedness as one of the dimensions of brand personality (Davies et al., 2001), which otherwise should not be there at all.Ruggedness is questioned in other studies also (Rojas-Mendez et al., 2004).Aaker's brand personality scale Aaker et al. (2001) found ruggedness to be replaced by peacefulness in case of Japan and Spain.The cultural differences were cited to be the reasons for this change.
New dimensions found by scholars worldwide, evidences of non-applicability of Aaker's dimensions and replacement of old dimensions with new dimensions should not be taken as surprise.Change in demographics, cultures, societies and lifestyles are bound to bring variation in results.Differences should rather be cherished, as it is a usual process of advancing any construct towards maturity.It is true that many studies are reporting different dimensions from Aaker but there are also studies that are reporting dimensions similar to that of Aaker's (Table IV).Differences and similarities are to be viewed from other perspective also.Differences can be attributed to the reasons cited above; similarities in dimensions can be viewed as support to the universality of the scale.Consider one of the potential scenarios: popularity is generally associated with magnetic effect.A popular work generally becomes base for others.The process adopted by original author is usually accepted as a standard process.A number of attempts on brand personality after Aaker are found to be following the process adopted by Aaker to arrive at dimensions of brand personality (most authors take the traits proposed by Aaker and apply factor analysis to find dimensions of brand personality).It is amazing to see how minutely success can be followed.For example, the profiling or naming of dimensions by Aaker as result of application of factor analysis might have great influence on how scholars would name the dimensions in their study.For example, in Aaker's original work, Factor 5 having outdoorsy, masculine, western, roughed and toughed traits is called as "ruggedness".The probability that authors across the globe, if happen to have these traits or say 70 per cent of these five traits clubbed under one factor, would call or name this factor as "ruggedness" might be quite high.While authors might not explicitly report it, but it might be a rule of thumb for them, while naming a factor.The phenomenon can surprisingly be true even in cases where authors might have another alternative name for those combinations of traits.This might explain the reasons for the similar dimensions reported by many scholars in their studies across cultures, nations, product and brand categories.

Methodology-related criticism
Methodology pertains to the method adopted to arrive at results.Aaker in her work first identified traits from different sources and filtered them down to manageable number.This approach is termed as quasilexical approach (identification of traits from different sources).This is one of the oldest and best methods to start the generation of items.Even the domain sampling theory used in questionnaire formulation recommend the generation of items from all possible sources (Churchill, 1979).Idea is to have an exhaustive list of all those items that could in anyway help in the measurement of the construct of interest.The exhaustive effort of Aaker to generate items is commendable (it still has objections, see the section "Conceptrelated criticism").
To find out dimensions of brand personality, factor analysis with varimax rotation was adopted by Aaker in the second stage.Usage of factor analysis to find out dimensions was not new when Aaker used it (for example, popular studies of Hofstede carried out from 1979 to 1983 to find out dimensions of cultureas brand personality and its dimensions are associated with human personality, dimensions of culture are related to psychological characteristics which relates to personalityused factor analytical approach) (Hofstede and Bond, 1984).Yet, a close scrutiny of Aaker's work depicts the objective for which she was working.As Aaker wrote in her article -"Because the objective of this stage was to identify the brand personality dimensions as perceived in consumers' minds [. ..]" (italics in original), it is clear that dimensions were to be reported as per consumers' perception.Here, objective was not only to reduce the data but also to find the latent dimensions in which consumers classify brand personality.In such cases, the use of principle component analysis is objectionable (Stewart et al., 2001;Smith et al., 2006).It is strongly argued that in such cases, the maximum likelihood (or other extraction methods of exploratory factor analysis) as extraction method with non-orthogonal rotation should be adopted to explore the dimensions.With the data of Aaker, variation in results with exploratory factor analysis can now only be imagined.Other similar questions relating to factor analysis have also been reported, such as: Sweeney and Brandon (2006) questioned the factor analytical approach to understand brand personality and argued to follow interpersonal circumplex (IPC) method as it is strong on theoretical ground and explains the interpersonal traits much better.Austin et al. (2003) strongly objected the use of factor analysis to draw framework of brand personality.Use of factor analysis-based model to find dimension of brand personality was also challenged by Heere (2010).Avis (2012) questioned the factor models for understanding brand personality.He charged Aaker (1997) and others by arguing that factor-based models generate the problems of category confusion (what is being measureda category or brand), domain adjustment (meaning of words changes when domain is changed) and descriptor selection.
Therefore, it is no surprise that other methods are advocated by scholars to study brand personality and related constructs.Arora and Stoner (2009) advocated the use of qualitative methods for measuring brand personality rather than quantitative methods.It is found that quantitative methods cannot capture in-depth detail of any product's personality.Study concluded that competence, sophistication and ruggedness dimensions are not coming out as brand personality dimensions in the qualitative approach of understanding brand personality.Hofstede et al. (2007) recommended using projective techniques to carry out brand image research rather than survey methods.Hence, study concluded personification to be a better method to study brand image.Still, scholars seem to have preferred the route of factor analytical approach as adopted by Aaker over other methodologies, such as Q-Methodology, IPC model, repertory grid analysis, psychological meaning and qualitative analysis (see Table V; from 28 studies reported in Table V, only five studies used a nonfactor analysis approach).This can be due to the popularity of Aaker's work.
Another strong argument challenging Aaker can be seen in the form of question on the idea of humanizing brands.Aaker is criticized for not delving into theory, as there are problems of theoretical support to it (Avis et al., 2012).Assigning human-like characteristics to brands is a-theoretical and has vague traces in history (Avis and Aitken, 2015).Following Aaker's methodology, even rocks (which were never promoted by marketer or anyone for personality associations) also have brand personality (Avis et al., 2013).It is because of the methodology that first assigns the personality (BPS starts with a note that makes the respondent to assume brand as human being and then respond) to brands and then report it.Hence, it can be called a mind game in which scholars are indirectly telling people to think of brands as humans.If one can imagine that, the brand personality is created which was never there (Avis et al., 2013).Huang et al. (2012) reported similar problem about BPS being atheoretical.BPSs in general are criticized for the confusion they create.Scholars even argued that BPSs are in fact not measuring brand personality for which they were formulated (Azoulay and Kapferer, 2003).
Another issue worth reporting relating to methodology is the level at which analysis is done.Aaker has carefully selected 37 brands to make the study representative of all brand categories.Ironically, presence of multiple brands in her study became a potential question.Cui et al. (2008).Variation in results for aggregate and individual level is not new to marketing literature.Literature is not void from the attempt to understand brand personality of individual brands using other techniques like Q-Methodology (Cui et al., 2008).It may be true that results may vary at individual brand level but then -Is there any other way to generate a scale that can be generalized across all categories of brands without taking a representative set of brands from various categories?Can we generalize the understanding of brand personality as derived from an individual brand?The answer is no!A representative set will be closer to generalization rather than an individual brand analysis.Moreover, no single study can be ultimate effort for any construct.The base Aaker generated can be refined further by subsequent studies.

Concept-related criticism
Despite all efforts of Aaker to make the trait list exhaustive, a list of 42 traits is called nonexhaustive and demand for addition of more traits to make the list comprehensive is witnessed (Sweeney and Brandon, 2006).The reason of this criticism can be found in the conceptual understanding of brand personality adopted by Aaker.To comment on this criticism, the process of generation and selection of traits by Aaker needs to be minutely analyzed.Churchill (1979) in his famous work documented the process of item generation and advocated it as an initial step.As part of this step, items should be comprehensively generated from the entire domain of construct (in this case, brand personality).The process (domain defining and item generation) itself starts from definition.But definition of brand personality adopted by Aaker (1997) has itself been questioned (see the section "Definitionrelated criticism").If the arguments against the way Aaker defined brand personality are substantial, then it can be called as a legitimate point from where the problem is originating.This might have given rise to the basic problem of conceptual completeness and validity (Carlson and Donavan, 2013;Heere, 2010) with BPS.Other problems highlighted by researcherssuch as BPS comprises only positive attributes of brands (Sweeney and Brandon, 2006;Bosnjak et al., 2007), BPS has created conceptual confusion (Azoulay and Kapferer, 2003;Heere, 2010), BPS does not give any conceptual support to five dimensions (Heere, 2010), BPS is not addressing the complexity of consumers interaction with brands (Smith et al., 2006) and BPS is crude measure of brand personality (Malik and Naeem, 2012) can be viewed as to be generating from ill-defined construct and domain specification.This is a more serious question, which should get the attention of researchers.If this is true, are we still waiting for the opening of the construct?
Conceptual problem can be cited as one of the root causes of another issue reported in "dimension-related criticism"the over dependence of BPS on big five model of human personality.The big five model should not be taken as a base for formulating or understanding brand personality without necessary changes in context of brand personality (Caprara et al., 2001).This leads to a legitimate problem of traits being adopted blindly from human personality literature and applied to brands (Azoulay and Kapferer, 2003;Sweeney and Brandon, 2006).Even, Aaker (1997) accepted and agreed that many human personality traits may not be applied to brands.Considering the adoption of human personality traits for brands without proper treatment may result in to a conceptual error, which Aaker is questioned for.Is this not taking the community back to where we have started?
This explains why Azoulay and Kapferer (2003) wrote the paper -"Do brand personality scales really measure brand personality?",Geuens et al. (2009) wrote -"A new measure of brand personality", Arora and Stoner (2009) wrote -"A mixed method approach to understanding brand personality", etc.The concept of brand personality is yet to be explored on these aspects.

Ethnocentric (words-related) criticism
Researchers have highlighted concern over the meaning of words used by Aaker when they are used in different culture and setting (Avis, 2012).One word/factor or trait when used for different brand may convey different meaning.Examples of such words are conscientious, stable and creative (Caprara et al., 2001).
Words that Aaker used in BPS were drawn in American context.This may lead to a potential problem when scale is implemented across countries.This might result in bringing the reliability of the scale down.The classic example of this is the traitwestern.Respondents might see "western" as an American or Non-Asian trait.If a respondent is asked to rate an Asian or say Indian brand on "western" trait, he/she may not relate the trait to such brand.Moreover, different traits may be interpreted differently for different brands by respondents (Austin et al., 2003).For example, respondents may not justify the usage of trait "rough" for baby products of Johnson & Johnson, but it may make sense in brand of bikes targeting adventurous customers.Such ambiguous traits might have managed to find place in BPS.Other problem creator words used for traits as highlighted by scholars are feminine, small town (Davies et al., 2001), upper class (Geuens et al., 2009), family oriented, outdoorsy and cool (Austin et al., 2003).It is very important that such traits should be used with caution in different setting as some of these may suit the brands in one specific area and some may not.
This ambiguity highlighted by researchers about certain traits in BPS should be carefully scrutinized before accepting the argument blindly.For example, the argument that trait "western" is belonging to America but not Asia and hence should not be used in non-American geographies need to be reviewed.The word "western" as trait should not blindly and etymologically be read as country specific as in scale it represents a lifestyle."Western" depicts boldness, adventurousness and rationality in life.Indians or Asians in general are not bold, adventurous and rational.Hence, they score low on "western".Scoring low on a trait has a meaning.Low score should not be read as non-applicability.The problem point in usage of traits like "western" in non-western countries is that the consumer may not view the trait "western" as a lifestyle.As Bishnoi and Kumar (2016) wrote in their paper: Considering the subject matter of study (accuracy of responses for BPS rests on the ability of respondent to assume a brand as a living being), which is relatively new, might require the researcher involvement in explaining questions to respondent [. ..] (Bishnoi and Kumar, 2016) (italics in original).
The argument here is that when BPS is given as a self-reporting questionnaire to respondent, probability that respondent would assign obvious meaning to the traits based on the word used to represent it is high.The remedy is involvement of researcher in explaining the true meaning of trait to respondent.Hence, scheduling would be a better way to record responses on BPS.If studies using BPS follow scheduling as data collection method, problems like these might be addressed.

Generalizability-related criticism
All the above criticism highlighted for BPS end up in challenging the universal applicability of scale.Therefore, it is not a surprise that scholars have reported issues with the generalizability of BPS (Austin et al., 2003).There are questions on BPS for its performance across product categories, cultures, and its generalization across languages (Das et al., 2012).Austin et al. (2003) highlighted the problem multiple times in their paper.The gravity of objections can be assessed from the arguments in article such as:  Austin et al. (2003) gave many reasons for "questioning the generalizability" of Aaker's scale, such as overlooking the concept of facets of differentiation and generalization, usage of factor analysis, BPS framework is derived for aggregate brands and not for individual brands, structural problems and non-defining of application areas.Pertaining to these reasons, researchers have developed new scales to overcome such problems.

Work of
As highlighted earlier, question on generalizability is not the core problem, but is the result of other problems associated with scale.Better way to address this problem would be to address the core problems rather than attempting generalizability.Moreover, in behavioural science such as marketing, generalization of construct in absolute terms is unattainable or impossible.

Conclusion
Use of metaphor in literature is prevalent but it should be used with caution.Metaphors help in understanding the complex role of brand.Acceptance of the metaphor converts it into a term (Davies and Chun, 2003).Brand personality has passed through the transition phase when it was called a metaphor.It is well accepted today as a term in marketing literature and industry.Aaker has taken the concept ahead by giving BPS.Scholars vehemently found themselves interested in taking it as a topic of research (Aaker's work has more than 7,559 google scholar citations as on July 2017).Many refused to accept Aaker's work and developed their own scales to measure brand personality pertaining to the issues highlighted in this paper (Table VI).
Criticisms highlighted here can be classified into two domains: (1) needing scholars' attention for refinement; and (2) attributable to evident reasons and hence not treated as potential threat.
Definition, conceptual understanding and methodology related criticism may seek the attention of scholars, as these are major allegations on Aaker.Moreover, these aspects are base of any construct.It is the responsibility of research community to filter down the suspected aspects related to brand personality.Criticism of dimensions, the type of words used along with generalizability, can be associated with culture, the way the data are collected (methodology) and meaning given to words by respondents because of their native language (these have been explained in the relevant sections of this paper).This variedness is natural and Aaker also accepted this in her article.It is true that Aaker's work has given a reason to scholars to work in this field.At the same time, it is interesting to note that Aaker herself has not published much on brand personality.Last paper of Aaker on brand personality was in 2005.Last paper that Aaker published on brand was in 2012 (Aaker et al., 2012).The reason why Aaker left the brand personality research is not known at least from the information available in public domain.She herself could have been the appropriate respondent to the questions raised on BPS.But now, the responsibility lies with the scholarly community to take it ahead.One aspect of her seminal research, which one has to agree with, is that despite all these criticisms, even researchers who criticized the BPS and all those who introduced new measures of brand personality have used directly or indirectly 42 traits of brand personality identified by Aaker (Table IV).With advent of BPS, different methods have been attempted by researchers resulting into different conclusions than that of Aaker.Aaker herself stated this as limitation of her work.Aaker accepted that the model she is proposing may give different results across different regions.Success of any construct and its measurement should not be looked in isolation.Any scale should be evaluated and analyzed for its effectiveness on parameters, such as ease of use, ability to give reliable results, validation and acceptance across the globe.Aaker's scale and work have been able to score high on these parameters.Literature has been witnessing that any massively popular work has invited other versions of it in the form of criticism, which is good for the development of the scope of construct.At least credit of motivating, enticing and sparking the researchers across the globe to work on brand personality can be given to Aaker, which in itself is a big contribution.

Scope for future work
It is time when research community should look at other aspects of brand personality rather than looking at the criticism of a specific scale, which started this debate.There are some important questions that need immediate attention from scholars, such as: (1) There is dearth of studies on legitimate aspects of brand personality, for example, anthropomorphism.
Consumers' ability to anthropomorphize is an important research topic that can contribute to the literature of brand personality.Relationship between consumers' liking, preference towards brands and ability to anthropomorphize are already attempted (Aggarwal and Mcgill, 2012)   (2) How brands acquire personality?What could be different modes of associations that could result into formulation of robust brand personality?(3) Innovations associated with brand and its role in formulation of brand personality should be addressed by scholars.(4) Research on impact of brand personality of a brand on the personality of humans who use it is in nascent stage.(5) Can a brand have two opposite brand personalities at the same time for different consumers?What could be the implications of it?

Table III .
It is argued that presence of multiple brands has made the study to give results at aggregate level.Same results might not have been generated at individual level (for single brand) as . ..]Unfortunately, it is not entirely clear in Aaker's article from what and to what the brand personality framework is generalizable [. ..] (italics in original).and [. ..] it is crucial to marketing thought and practice that the boundary conditions for the generalizability of Aaker's research conclusions first be identified [. ..] (italics in original).and [. ..] but rather to make clear the need to submit the framework to additional testing before concluding it is 'generalizable' for measuring the personality of individual brands [. ..] (italics in original). [ but these are very few.What can be anthropomorphize and what cannot is another important research area.