‘No other reviews, no purchase, no wish list’: Book reviews and community norms on Amazon.com

Anni Sairio
University of Helsinki

Abstract

This article presents a discourse-analytical case study of a book review and the responses it has attracted on Amazon.com in order to explore aspects of trust and group norms in this experience-sharing, trust-based commercial online community. The reception of this highly ranked review suggests that there are contrasting norms at work: some commenters emphasize the use of standard language as a measure of quality and look to the reviewer’s activity as a measure of trust, while others consider a reader’s subjective contribution of their reading experience to be the key element. User regulations of Amazon.com direct members to operate primarily on the basis of identity-based attachments, but reviewing activities themselves seem to reflect both identity-based and bond-based attachments to the community. The participants’ disagreement with regard to norms and ideal practices perhaps mark the Amazon.com reviewing community to be in the early stages of community development (see Wenger et al. 2002).

When it comes to assessing the validity of the reviewing process, an important tip is to beware the extremes − reviews that are chock-full of exclamation marks praising the work of the unknown author, yet written by people who’ve never written any other reviews in the past, as well as nasty 1-star reviews that are unnecessarily brutal or nit-picking.
– Elisa Hategan, Alice in Writerland (2012: 340)

You’re criticizing people who’ve been here for years when you have a blank profile and are accusing others of being trolls. I’ve got at least a dozen pages of products I’ve reviewed, going back at least to 2004. You have-----Oh, nothing.
– username ‘ginmar’ on Amazon.com (May 2012)

1. Introduction

Book reviews on Amazon.com have caused controversy in recent years on account of sockpuppet incidents and other attempts to game the system. Sockpuppetry (or sock-puppeting) has been defined in the New York Times as “the act of creating a fake online identity to praise, defend or create the illusion of support for one’s self, allies or company”, and cases involving historians and novelists have recently come to light (see the 2010 Guardian article on historian Orlando Figes who posted reviews under an alias trashing his rivals and the 2012 Huffington Post overview on the incident surrounding crime novelist R. J. Ellory). The emerging book review industry enables authors to purchase reviews, and in 2013 a fan campaign attempted to hurt the sales of a Michael Jackson biography by bombarding Amazon.com with one-star reviews. As an online retailer, Amazon.com has an interest in ensuring and maintaining the reliability of customer reviews, which are shown to have a positive influence on sales especially if the reviews are very positive (Clemons et al. 2006, Chen, Dhanasobhon and Smith 2001, Mudambi and Schoff 2010). They have responded to these incidents by removing reviews and placing certain restrictions, an approach which so far has not seemed to solve the problems.

Wiertz & Ruyter (2007: 349) define commercial online communities as “firm-hosted online aggregations of customers who collectively co-produce and consume content about a commercial activity that is central to their interest by exchanging intangible resources”. Wasko-McLure & Faraj (2000) have shown that individuals participate in computer-related online communities in order to obtain information, to learn, and to reciprocate the assistance they have previously received; individuals help strangers also to enhance their professional reputations and, to some extent, because it is enjoyable (Wasko-McLure & Faraj 2005). User experiences are a key element in the function of the reviews, and these experience-related interactions on Amazon.com are based on trust (see Kim & Phalak 2012). For the retailer, customer reviews provide reliable user experiences and encourage sales by assisting in potential purchasing decisions. Book reviews are particularly important for self-published authors, as they are considered to “stabilize[...] a writer’s authority” (Laquintano 2010: 486). Interestingly, book reviews are considered more helpful when they are moderate rather than extreme (Forman, Ghose & Wiesenfeld 2008). Book reviewers engage in order to contribute to literary discussion and criticism, to exchange and sound out opinions and reader experiences, to seek information, to share experiences, and to connect with other readers and, perhaps, with the author: “Writing a review is, second only to purchase, possibly the best way I can say ‘thank you’ to the author for the effort and energy expended,” explains book blogger I am, Indeed. Reviews are written also to amuse (on a t-shirt printed with the image of three wolves howling at the moon: “Unfortunately I already had this exact picture tattooed on my chest, but this shirt is very useful in colder weather”). And although it breaks the regulations, some reviewers may take part for their own business purposes, to boost sales with glowing reviews or to attempt the opposite. These various motivations result in what seems to be a heterogeneous online community with potentially conflicting dimensions.

Amazon.com is a commercial online community where significant changes currently unfolding in institutional publishing practices are negotiated among writers, readers, and reviewers, and this paper presents a discourse-analytical case study of a book review and the response it has garnered in order to explore community aspects of trust and norms on Amazon.com and to outline potential research angles in this vast range of publicly available, both text-based and multimodal “peer-generated product evaluations” (Mudambi & Schoff 2010: 186). Specifically, what do the responses to the review convey of participant perceptions as to what constitutes a proper review? How do presumed network ties and standard language ideologies influence the participants’ evaluation of the review? And does this interaction within the literature space of Amazon.com appear to constitute a community with a set of shared norms and a shared purpose, which would seem to be ideal in terms of trust-building? Given the complexities that surround the reviews, including the motives of the reviewers and the participants, can we talk of shared community norms? The selected review, titled ‘WOW!!! I am a BELIEVER!!!!’, is a relatively short evaluation of a best-selling self-published novel from 2011, which other participants have given a high numerical rating. The review contains a considerable amount of positive affect markers (Biber & Finegan 1989) and it focuses intently on the reading experience: the actual contents of the book are not discussed. I have selected this review because of its content characteristics, the considerable attention it has attracted among other reviewers and readers, and its focus on a self-published novel, an increasingly noteworthy form of publication which continues to generate debate due to the lack of gatekeepers. Possible community norms and ideals are investigated particularly through stance-taking, and questions of identity emerge as a key element in assessing trust.

The Amazon.com reviews and responses are publicly available as they are intended to be perused by a wide audience, so the data is treated as public domain material. However, social media presents considerable ethical concerns for scholars, and there is without a doubt a difference between posting information to be publicly available and having it used for a purpose other than originally intended (D’Arcy & Young 2012). Individuals may invest in their online identities to a high degree, and researchers should consider whether the same kind of confidentiality should be given to these personae as in real life (see, for example, Frankel & Siang 1999: 17). An attempt has been made to focus on the contents of the interaction rather than the individuals involved.

2. Characteristics of virtual communities

First I will discuss the defining characteristics of virtual communities, and then describe the reviewing practices on Amazon.com to illustrate the make-up and the encouraged conventions of this vast online space. Herring (2008) distinguishes the three broadly used meanings of virtual community as, first of all, the social behaviour of a group in a computer-mediated environment, the phenomenon of online group formation, and the technological environment behind online group formation and activity. Herring (2008) points out that the term is often used aspirationally, without objective evaluation of community characteristics. Hummel & Lechnel’s (2002) four dimensions of community are 1) a clearly defined group, 2) interaction between members, 3) bonding among members, and 4) a common place (see Steinfield 2004: 222) for an adaptation of their model), while commonly-accepted criteria for virtual communities include 1) a shared reason for communicating, 2) the existence of norms and protocols, and 3) regular interaction over time that takes place in the internet, through a common mechanism (Herring 2008). Hummel & Lechnel’s (2002) community characteristics can be viewed as integrated into the dimensions of virtual communities. Voluntary commitment, participation, and contributions form the basis for almost all online communities (Ren, Kraut & Kiesler 2007). If we compare communities of practice (CoP, collaborative and task-related community spaces) with virtual communities of practice (VCoP), a key difference is the technological component of VCoPs which results in different environments and may influence the sense of belonging, mutual knowledge, and trust among members (see Dubé, Bourhis & Jacob 2006: 70).

According to Ganley & Lampe (2009), virtual communities differ from traditional communities in that (for one thing) they may not have a predefined structure to guide the relationships, which consequently results in increased openness, less risk for participants, and less information available about the participants. Kim & Ahmad (2013: 438, 449) point out that the challenge in collaborative, unregulated online communities is that the users are responsible for evaluating and validating the quality of content and the reliability of content providers. Trust is therefore a key issue in interactions among users (Kim & Phalak 2012). Trust is based on positive expectations of a satisfying outcome and the willingness to depend on another, whereas distrust is governed by negative emotions and insecurity regarding a user’s motivation, intention, and behaviour (McKnight & Chervany 2001, quoted in Kim & Ahmad 2013). Mishler & Rose (1997) distinguish lack of confidence from distrust as a lower level of trust or scepticism. Recommender and reputation systems in online markets aid in building trust due to the history of interactions and the explicit, experience-based feedback they provide (Terveen & Hill 2002, Dellarocas 2003, Zhou, Dresner & Windle 2008, Kim & Phalak 2012).

The review system on Amazon.com is clearly constructed as community-based activity. The members are required to sign up if they wish to post reviews and messages, and they then agree to follow the retailer’s rules of conduct; as Hummel & Lechnel (2002) point out, the existence of rules is obviously not the same as their implementation. The discussion forums and the possibility to comment on the reviews enable interaction between members. Community bonding manifests itself in friendships, trust, and identification of members (Steinfield 2004: 222), and this is to some extent enabled through viewing user profiles (including reviews and users’ overall rating as reviewers) and interaction on the discussion forums and the reviews themselves, which are two distinct spaces as will be illustrated. The focus of the community is determined by a common language, common values, common interest, and the motivation to contribute (Hummel & Lechnel 2002). Could we assume reviewing activities to display shared values, a sense of belonging and a joint purpose to the degree that the participants actually form a community? A sample of one may point towards a hypothesis of whether or not the participants share such a purpose and even form a community of practice (Wenger 1998, Holmes & Meyerhoff 1999). If we consider reviewing as the activity of a community of practice, then this VCofP would be based upon the sharing of user experiences to provide information for potential readers and a space for reading-related interaction. This potential VCoP would build upon trust and constantly evaluate the reliability of participants. As the motives of participation vary, there may be different sets of norms and ideals as well as purposes to engage in this community. These aspects will be discussed further in this paper.

Individuals’ attachments to online communities can be explored through the concepts of identity-based and bond-based attachments (Ren, Kraut & Kiesler 2007). Identity-based attachments imply commitment to and focus on the purpose of the community (which presupposes that participants agree what the purpose is), whereas bond-based attachments result from social or emotional ties to other members and rely on the development of friendships to maintain the coherence of the group (Ren, Kraut & Kiesler 2007: 381, 390), which may, at least to some extent, be realized independent of the group’s purpose. Identity-based groups display high conformity to group norms, hold on-topic discussion in great value, and show low tolerance towards off-topic conversations, all factors which distinguish them from bond-based groups (2007: 392). From the retailer’s point of view, identity-based attachments would seem to be a desirable foundation for a reviewing community. The Amazon.com guidelines presented in the following section indicate that this is so, but the interaction itself suggests that both types of attachment are present at least in certain contexts.

2.1 Characteristics of reviewing practices on Amazon.com

Amazon.com can be characterised as an experience-sharing, trust-based commercial online community with large-scale pseudonymous participation (see Kim & Phalak 2012, Wiertz & Ruyter 2007). Participation is regulated. A password-protected account is required to create a profile page, in which participants can set up a username, choose an avatar, and post a brief introduction of themselves. The profile page contains a list of reviews the participants have written and possibly a wish list of future purchases. This publicly accessible information contributes toward creating an online identity, but Amazon.com users may also be assigned identities on the basis of their activities, a process which to some extent is out of their hands. Registered users are able evaluate reviews as helpful or unhelpful (how these two concepts are understood and this function put to use could be discussed). The ‘Top Reviewer’ status is achieved on the basis of these votes and the number of contributions, so the helpful/unhelpful votes influence a user’s reliability. A sufficient number of ‘unhelpful’ votes in the review comments can result in this comment being hidden.

The Amazon guidelines instruct reviewers to explain why they have or have not enjoyed the product in question, to be specific and sincere in their review, to adhere to word count limitations (an ideal review contains 75 to 500 words, and a video review should be between 2 to 5 minutes), and to disclose clearly if the product has been received in exchange for a review. The scale of the numerical evaluation ranges from one to five stars. The company reserves the right to remove reviews which contain promotional content (including reviews written “for any form of compensation other than a free copy of the product”), objectionable material (“obscene or distasteful content, profanity or spiteful remarks, promotion of illegal or immoral conduct”), inappropriate content (including excessive quoting and use of foreign language without a clear contextual reason), and off-topic information (shipping information, alternative ordering, typos in the product description). The retailer thus directs the members to operate primarily on the basis of identity-based attachments.

The reviews can be posted as simple text or in multimodal form. The reviews are posted on the product site, and the interaction is restricted to a review−response pattern (see Figure 1). To see the users’ avatars, one must visit their profile page. In this respect the reviews differ from more clearly community-oriented spaces, such as Amazon’s own discussion forums (Figure 2), which provide a place for more wide-range discussion and interaction.

Figure 1. Product reviews on Amazon.com. (Source)

Figure 2. A discussion forum on Amazon.com (Source)

Dubé, Bourhis & Jacob’s (2006: 70) comment about virtual communities of practice that are created or mainly used by organizations may be relevant in this context:

[...] a VCoP created in an organization may be highly disruptive and stressful for the people involved. As an organizational innovation – and even more so if participation becomes mandatory – it may force people into unfamiliar roles and ways of communicating/sharing, and into developing skills to create social ties through technology.

Participation and norms of behaviour on Amazon.com are, perhaps, not enforced to this extent, but compared to non-commercial online communities, the platforms, purposes of engagement, and social practices of which are decided and constructed by individuals rather than by organisations or corporations, interaction on Amazon.com is controlled or guided by practices established top-down by the organisation.

3. A review of the Mill River Recluse (2011)

The selected review evaluates an e-book titled the Mill River Recluse. The novel was originally self-published by Darcie Chan as a Kindle edition in the spring of 2011, and username Karen Grigg reviewed it in July of that year. Due to the success of the novel (spurred by the low price of $0.99 when it was first published), it was later picked up by Ballantine Books and published as a paperback in November 2012. Amazon.com lists it under general fiction, and it is characterized as “a good escape” and “a good beach read” on another review site.

The aspect of self-publishing is a key issue. An avalanche of self-published e-books began in the late 2000s with the launch of CreateSpace and Kindle Direct Publishing of Amazon.com, which allowed authors to upload their work directly to the Kindle store (Bradley, Fulton & Helm 2012: 135). Dilevko & Dali (2006) provide an overview of the developments in self-publishing from the late 1990s to the mid-2000s. Self-publishing has been regarded with misgivings due to the lack of gatekeepers to screen the quality of work and separate the wheat from the chaff, but the stigma appears to be disappearing (Dilevko and Dali 2006; see also the recent articles in Mediashift and Techdirt). Reviews are an essential part of promotion and marketing in the self-publishing industry (see Laquintano 2010).

In January 2013, there were 1,153 reviews of the Mill River Recluse on Amazon.com. Karen Grigg published hers on July 6 2011, not long after the self-published Kindle edition had come out, and it is ranked the most helpful by 1,107 (93%) of 1,185 voters. The review received 44 comments between July 22 2011 and January 15 2012; the discussion has now dwindled down, but users continue to rate the review since the number of votes has still continued to increase. The second most helpful review is rated as such by 448 of 478 users (January 2013), and there are three comments; this illustrates the popularity of Karen Grigg’s review. The review appears below, as accessed on January 2013:

Figure 3. (Source)

The review provides a descriptive narrative of the reading process, from the accidental discovery of the book (found thanks to the price and positive reviews) to the emotional and rewarding reading experience (laughter and tears). The reviewer has identified with the novel and its world, and she addresses both the author (“Well done!!!”) and the readers of the review (“Enjoy this read!!!”) in a positive and rather energetic style. Stancetaking is a complex and broad framework of linguistic analysis, and in this paper I apply Conrad & Biber’s (2000) and Biber & Finegan’s (1989) stance analysis. Attitudinal stance conveys speaker attitudes, feelings, or value judgements, epistemic stance conveys the certainty, reliability, or limitations of a proposition, and style stance describes the manner of presentation (Conrad & Biber 2000: 57). The positive affect markers (wonderfully surprised, enjoyed) and emphatics (really, so, just, capital letters, exclamation marks) categorize Karen Grigg’s review as affective, and it lacks evidentiality markers which would provide a more analytical approach with e.g. the use certainty and doubt verbs, adverbs, and adjectives (Biber & Finegan 1989: 116, 119–122); “I felt I HAD to write a review” includes a doubt verb, but feel in this context conveys emotion.

The responses to the review fall roughly into two categories: those that support and thank the reviewer for sharing their experience, and those that express doubt and criticism as to its purpose and execution. The writer of the review has no other activity on Amazon.com: to some, this is a crucial aspect, to others, insignificant. The first comments are made in a positive note, and they allude to the influence of reviews (“Thx for such an enthusiastic review. I think I will try this one”, “Your review made me want to read this book. Thank you so much”) and offer good-humoured advice on spelling (“Btw, ‘definitely’ does not have an ‘A’ in it :)”) amended with face-saving tactics (“just FYI, a common mistake by reviewers”). At first the discussion centres on the topic of spelling, and the supportive participants talk among themselves about typos and misspellings, express good-humoured embarrassment over their own spelling errors, return to edit their posts for misspelled words, and make ironic and disapproving remarks when others begin to chime in in a negative tone. The first negative comment (#8) “I’m going to read this book because of all the stupid REVIEWS! Caps not proper English” is met with responses such as “Sorry Karen for all the haters...this is a book review on Amazon not an English paper”, “if you guys want to teach an English class, please, get off of Amazon and do so. It’s a review, people”, and “I hope I spelled all that correctly!”. Gracie posts an ironic comment which perhaps makes fun of the spelling police, perhaps targets the spelling reviewers and like-minded enthusiasts: “I definAtely agree. And eye m going two bye this book and reed it now”. There is evident disagreement as to the expected correctness of language.

In addition to spelling, the negative responses are triggered by the reviewer’s lack of activity which indicates that credibility is directly linked to one’s status as a community member – though not for all participants. The supporters are not able to persuade the sceptics, as the following discussion shows:

(1) No other reviews. No purchase. No wish list. Not a purchase.
I wonder..... [ginmar, December 13, 2011, comment #38]
 
Would you have written any more reviews after the vicious attacks on your first one if it was like this person’s? [S. Warfield, December 13, 2011, comment #39]
 
If this is your idea of vicious (and for such a suspicious review) then I don’t know whether to envy or pity you. Either way, the writer’s friends need to recognize that if they flood her reviews, they’ll get nailed for it. [ginmar, December 14, 2011, comment #40. Emphasis added.]

Username ginmar (117 reviews in January 2013) identifies the reviewer as not a genuine member of the community but as having a vested interest to support the (at the time) self-published author. ‘WOW!!!’ is not perceived as an objective critique, and this suspected network tie between the reviewer and the author puts the reviewer’s reliability in questionable light. Amazon.com has attempted to find solutions to the problems of sockpuppetry, and the supportive participants are well aware of the meanings assigned to potential network ties with authors: some of them have reviewed this book themselves, and maintain their objectivity by including disclaimers such as “I am not in any way related, nor do I know the writer” (review of the Mill River Recluse titled ‘interesting and intriguing’, September 23, 2011).

Nevertheless, the supporters consider the subjectivity of the reading experience to be a key element of a book review: they respond positively to the reviewer’s enthusiasm, and indicate that it has a direct influence in their purchase decisions (“I love enthusiastic reviews like yours”; “I bought the book based on what you said”; “I make most of my decisions on what to buy based on customer reviews”). Active members, who have later posted their own reviews of the Mill River Recluse, encourage Karen Grigg to write more reviews. The expressions of distrust do not seem to sway them. A member with 667 reviews and a helpfulness rating of 92% writes:

(2) The person who wrote the original review has been treated so badly and unfairly. Maybe she enjoyed the book so much that she wanted to share that with others. I’m sure there isn’t one who has commented, including me, who hasn’t misspelled a word either here or somewhere else. Pointing out spelling errors is not good forum protocol. After this, the person won’t post reviews for sure. [S. Warfield, September 29, 2011, comment #22. Emphasis added.]

An established member comments here on appropriate community practices, and the majority of the participants clearly agree. A few comments earlier, a response by an almost equally established reviewer with 492 reviews (82% helpfulness rating) has been hidden as not adding to the discussion. This hidden comment (presented below) again categorically questions the reliability of the review on the basis of a suspected network tie to the author, implied in the lack of other presence in the community, but the majority of participants do not agree with this position.

Figure 4. (Source)

The practice of hiding comments is an interesting aspect of virtual communication, and in this thread altogether six comments have been hidden due to participant evaluation. These include snide and ironic comments about spelling errors (“Karen, you ‘DEFINATELY’ need to consult a dictionary before writing your next review”, “It is pathetic when reviews have misspelled words”), the above-mentioned sceptic remark (“I never believe in one-review reviewers”), cat-fight innuendo (“Couldn’t resist? MEEEEOW! Let’s all scratch each others’ eyes out”), and an off-topic comment on the proposed spelling reform of definitely (“Why not? But the “a” should be after the “f”, not the “n”. Or do we need a referendum on this issue?”). Curiously, one of the hidden comments is a critical take on the contents of the review and the lack of description:

(3) What was the book about, who were the characters? What made this book so good that you couldn’t put it down, make ME want to read this book. [Juli DeMas, October 9, 2011, comment # 27 − hidden]

The guidelines specifically encourage reviewers to provide this kind of information. Perhaps this comment is viewed as face-threatening and potentially discouraging for a new member. Comments made around the same time suggest that positive affective stance is held in high value:

(4) Thank you for your review of the writing & not a summary of the story. No spoilers here! [...] [Librarian, October 5, 2011, comment #25]
 
[...] Karen − I really enjoyed your review. As someone else mentioned, it’s nice to get a review of the book without getting a blow-by-blow plot summary. I can get that from Amazon. Keep up the reviews and ignore the “perfect people.” :) [S. Garrett, October 11, 2011, comment #29]
 
Re: S. Garrett --- This is what I think!! The readers are getting the “blow by blow plot summaries” from Amazon, so I try to find other ways to explain why I like the book. [RosaSharon, October 11, 2011, comment #30]

A subjective reading experience conveyed to other participants with the means (if not the intention) to engage them in a discussion appears to be considerably more meaningful and valuable than critical analysis or information about the story, which these participants expect Amazon.com to provide in the product description – product descriptions explain the contents, while the reviews contain the experiences. The discussion dwindles down after a content-related comment which addresses the main character’s agoraphobia and anxiety disorder: “I think the book sends a terrible message of not dealing with problems and others enabling the avoidance techniques,” writes username Sage in January 15, 2012 in what is currently the last comment of the thread. Creating and maintaining a positive tone of discussion and a supportive atmosphere for new members is evidently considered more important than trust-related evaluation of the reviewer’s motives and competence. The reviewer does not return to comment, so we do not know their stance on the discussion that follows.

4. Discussion

The reception of this review suggests that the participants construct conflicting norms. For some, the language of the review and the reviewer’s previous activities are key criteria, and when these conditions are not satisfactorily met, they invalidate the entire review. Others find this irrelevant. Perhaps the numerical rating provides the information they seek, and the positive and affective stance of the review outweighs any technical details. They consider the reviewer’s subjective contribution to be the key element, and the actual execution is of secondary importance. Perhaps the genre of the novel is an important variable. This is evidently a comfort read, and it is possible that readers want to retain that same feeling of warmth and comfort in the discussion. Reviews of different genres and different products could well provide different results.

Irvine’s (1989: 255) understanding of language ideologies emphasises the sociocultural aspect of linguistic evaluations and their “loading of moral and political interests”. Standard language ideology (Lippi-Green 1997, Milroy & Milroy 1999, Kroskrity 2004) is behind the many comments on spelling errors (ridicule, dismissing the review, also the positive corrections). Defined as “a bias toward an abstracted, idealized, homogenous spoken language which is imposed and maintained by dominant bloc institutions and which names as its model the written language, but which is drawn primarily from the speech of the upper, middle class” (Lippi-Green 1997: 64), standard language ideology helps to explain why the review is dismissed partly on account of language. The comments on reforming the spelling of definitely (“I would be willing to start a campaign to change the spelling to definately. Definitely must have been started in the first place by someone hoping to trip all of us up!”, #6) can also be read as light-hearted contesting of a dominant ideology. Irvine and Gal (2000: 38–39) refer to the erasure of unruly variation as the process in which “ideology, in simplifying the sociolinguistic field, renders some persons or activities (or sociolinguistic phenomena) invisible” – an opinion is invalidated on account of imperfect execution. Nevertheless, this attitude is challenged by the majority of participants, including the large crowd of silent individuals whose presence is detected only in the number of votes Karen Grigg’s review has received. We may assume that lurkers or passive readers far outnumber those who post comments (Preece, Nonnecke & Andrews 2004, Schlosser 2005) and the high number of votes suggests that Karen Grigg’s review is considered in a generally positive light, be that ironically or otherwise. Her lack of history on Amazon.com, which in the sceptics’ minds implies that she has created an account merely to write a favourable review to boost the sales, does not appear to matter to most participants, four of whom will eventually review the Mill River Recluse themselves and clearly look to engage in discussion about the book.

According to Ren, Kraut & Kiesler (2007: 390), “online community members who feel identity-based attachment will be more likely to conform to group norms than those who feel bond-based attachment to the community”. The criticism of spelling suggests that these members show identity-based attachment to their ideal of the community, and standard language is evidently part of the norms that need to be met in order for a participant or a comment to be validated. This is, however, the minority view. Other participants put more weight on the shared experience, state that their decision to purchase the novel spans directly from the review, and criticise those who do not contribute to the positive atmosphere. This stance is evident also in the practice of hiding critical comments. There seems to be discrepancy in the norms that are displayed: the significance of the reviewer’s activity in the community, language ideologies, and the subjective reader experience all have different meanings and values to participants. Given the variety of motivations for taking part, this discrepancy is not surprising. Reviews may be written as genuine evaluations, for amusement, and as illicit sockpuppet activities and they are also read for different reasons, so any community norms are likely to depend on a number of variables despite the set of guidelines that theoretically function as community rules. Since the participants do not agree on the norms and protocols and the purpose of communication, it seems that we cannot define this community as a virtual community of practice; it is a community, given that the participants share an overall purpose of producing user experiences and they do this within a specific, clearly bounded, and shared online environment, but they cannot reliably identify each other and agree on their ideal forms of practice. Perhaps the reviewing community on Amazon.com has not yet reached that stage of development where the members agree on the joint purpose of the group; see Wenger et al. (2002: 69) on the five stages of community development from potential to transformation.

This case study indicates that online reviewing practices offer considerable potential for network analysis and quantitative and qualitative CMC research in terms of in-depth and long-term observation and the (perhaps limited) use of ethnographic methods (see Langman & Sayer 2013, D’Arcy & Young 2012). This is a small study that points to the possibilities of big data approach, for example with regard to possible patterns in hidden comments and the correlation between the number of reviews and helpfulness rating and the perceptions of group norms and review ideals. And large-scale comparative analysis of, for example, the reception of five-star reviews of self-published and traditionally published novels of the same genre could provide more insights into this relatively recent turmoil in the publishing world.

5. Conclusion

A five-star review of a self-published novel on Amazon.com receives positive responses for its affective and personal stance, and criticism due to its informal execution and the reviewer’s lack of other activity, a factor which brands this newcomer as unreliable. This distrust is expressed against the backdrop of recent events where individuals ranging from devoted fans to published authors have attempted to influence sales and product ratings by posting reviews under fake online identities. The majority of participants do not find these concerns to be of much substance, however – for most, a positive reading experience of a feel-good novel appears to be the key issue. On the basis of this small case study, users seem to have contrasting expectations as to what constitutes a proper book review, and the Amazon.com guidelines do not function as a measuring stick. This interaction indicates that the community members disagree as to the community norms and the ideals of a proper review. The genre of the reviewed work and the publication path may be key variables, given how important reviews are for the promotion of self-published books. Big data analysis would provide insight into whether the responses to Karen Grigg’s review of the Mill River Recluse make an exception or whether the community norms and ideal practices are indeed more or less heterogeneous by nature, thus perhaps marking the Amazon.com reviewing community to be in the early stages of community development.

Sources

New York Times definition of ‘sock puppetry’: http://www.nytimes.com/2007/07/16/technology/16blog.html

Guardian article on Orlando Figes: https://www.theguardian.com/books/2010/apr/23/historian-orlando-figes-amazon-reviews-rivals

Huffington Post article on R. J. Ellory: http://www.huffingtonpost.com/2012/09/04/rj-ellory-fake-amazon-reviews-caught_n_1854713.html

New York Times article on the ability of authors to purchase online reviews: http://www.nytimes.com/2012/08/26/business/book-reviewers-for-hire-meet-a-demand-for-online-raves.html

New York Times article on the attempt to hurt the sales of a Michael Jackson biography: http://www.nytimes.com/2013/01/21/business/a-casualty-on-the-battlefield-of-amazons-partisan-book-reviews.html

Guardian article about Amazon.com’s attempt to solve the problems with its review system: https://www.theguardian.com/books/booksblog/2013/jan/25/why-amazon-just-cant-win

Book blogger I am, Indeed on reviewing: http://iam-indeed.com/why-i-write-book-reviews/

Wolf t-shirt: https://www.buzzfeed.com/annanorth/the-26-best-amazon-reviews-of-all-time?sub=1818698_638127&utm_term=.saV7DVBPX#.jsZOpJNoD

Amazon guidelines: https://www.amazon.com/gp/help/customer/display.html?nodeId=201929730

The Mill River Recluse review on bestsellers.about.com: http://bestsellers.about.com/od/fictionreviews/fr/The-Mill-River-Recluse-By-Darcie-Chan-Book-Review.htm

Mediashift article on self-publishing: http://mediashift.org/2010/12/2010-the-year-self-publishing-lost-its-stigma363/

Techdirt article on self-publishing: https://www.techdirt.com/articles/20120910/05584620327/ebook-authors-continue-to-see-self-publishing-stigma-disappear.shtml

Source for Figure 1 on Amazon.com: https://www.amazon.com/Darkness-Sethanon-Riftwar-Saga/dp/0007229437/

Source for Figure 2 on Amazon.com: https://www.amazon.com/forum/top%20reviewers

Source for Figure 3 on Amazon.com: https://www.amazon.com/review/R1B8Z78LL13VOI/

Source for Figure 4 on Amazon.com: https://www.amazon.com/review/R1B8Z78LL13VOI/ref=cm_cr_rdp_perm?ie=UTF8&ASIN=B0051PRFLQ

References

Biber, Douglas & Edward Finegan. 1989. “Styles of stance in English: Lexical and grammatical marking of evidentiality and affect”. Text & Talk 9(1). 93–124.

Bradley, Jana, Bruce Fulton & Marlene Helm. 2012. “Self-published books: An empirical ‘snapshot’”. The Library Quarterly 82(2): 107–140.

Chen, Pei-yu, Samita Dhanasobhon & Michael D. Smith. 2001. “All reviews are not created equal: the disaggregate impact of reviews on sales on Amazon.com”. Heinz Research. Paper 55. http://repository.cmu.edu/cgi/viewcontent.cgi?article=1054&context=heinzworks

Clemons, E., G. Gao & L. Hitt. 2006. “When online reviews meet hyperdifferentiation: A study of the craft beer industry”. Journal of Management Information Systems 23(2): 149–171.

Conrad, Susan & Douglas Biber. 2000. “Adverbial marking of stance in speech and writing”. Evaluation in Text. Authorial Stance and the Construction of Discourse, ed. by S. Hunston & G. Thompson, 56–73. Oxford: Oxford University Press.

D’Arcy, Alexandra & Taylor Marie Young. 2012. “Ethics and social media: Implications for sociolinguistics in the networked public”. Journal of Sociolinguistics 16(4): 532–546.

Dawson, Laura. 2008. “The role of self-publishing in libraries”. Library Trends 57(1): 43–51.

Dellarocas, Crysanthos. 2003. “The digitization of word of mouth: Promise and challenges of online feedback mechanisms”. Management Science 49(10): 1407–1424.

Dilevko, Juris & Keren Dali. 2006. “The self-publishing phenomenon and libraries”. Library and Information Science Research 28(2): 208–34.

Dubé, Line, Anne Bourhis & Réal Jacob. 2006. “Towards a typology of Virtual Communities of Practice”. Interdisciplinary Journal of Information, Knowledge, and Management 1: 69–92.

Forman, Chris, Anindya Ghose & Batia Wiesenfeld. 2008. “Examining the relationship between reviews and sales: The role of reviewer identity disclosure in electronic markets”. Information Systems Research 19(3): 291–313.

Frankel, Mark S. & Sanyin Siang. 1999. “Ethical and legal aspects of human subjects research on the Internet”. Report of a workshop, June 10–11, 1999. Washington DC. http://www.aaas.org/sites/default/files/migrate/uploads/report_hsr.pdf

Ganley, Dale & Cliff Lampe. 2009. “The ties that bind: Social network principles in online communities”. Decision Support Systems 47(3): 266–274. http://gsb.haifa.ac.il/~draban/SocialDynamics/ties.pdf

Herring, Susan. 2008. “Virtual community”. Encyclopedia of Qualitative Research Methods ed. by Lisa M. Given, 920–921. Thousand Oaks, CA: Sage. Preprint: http://ella.slis.indiana.edu/~herring/vc.pdf

Holmes, Janet & Miriam Meyerhoff. 1999. “The Community of practice: Theories and methodologies in language and gender research”. Language in Society 28: 173–183.

Hummel, Johannes & Ulrike Lechner. 2002. “Social profiles of virtual communities”. Proceedings of the 35th Hawaii International Conference on System Sciences 7, 172b. IEEE Computer Society. doi:https://www.computer.org/csdl/proceedings/hicss/2002/1435/07/14350172b-abs.html

Irvine, Judith. 1989. “When talk isn’t cheap: Language and political economy”. American Ethnologist 16: 248–267.

Irvine, Judith & Susan Gal. 2000. “Language ideology and linguistic differentation”. Regimes of Language, ed. by Paul Kroskrity, 35–83. Santa Fe, NM: School of American Research Press.

Kim, Young Ae & Muhammad A. Ahmad. 2013. “Trust, distrust and lack of confidence of users in online social media-sharing communities”. Knowledge-Based Systems 37: 438–450.

Kim, Young Ae & Rasik Phalak. 2012. “A trust prediction framework in rating-based experience sharing social networks without a Web of Trust”. Information Sciences 191: 128–145.

Kroskrity, Paul V. 2004. “Language ideologies”. A Companion to Linguistic Anthropology, ed. by Alessandro Duranti, 497–517. Malden, MA and Oxford: Blackwell.

Langman, Juliet & Peter Sayer. 2013. “Qualitative sociolinguistics research”. The Encyclopedia of Applied Linguistics, ed. by Carol A. Chapelle. Oxford: Wiley-Blackwell. doi:10.1002/9781405198431.wbeal0985

Laquintano, Tim. 2010. “Sustained authorship: Digital writing, self-publishing, and the e-book”. Written Communication 27(4): 469–493.

Lippi-Green, Rosina. 1997. English With An Accent: Language, Ideology, and Discrimination in the United States. London: Routledge.

McKnight, Harrison D. & Norman L. Chervany. 2001. “Trust and distrust definitions: One bite at a time”. Trust in Cyber-Societies, ed. by R. Falcone, M. Singh, & Y.-H. Tan, 27–54. Berlin: Springer-Verlag. http://link.springer.com/chapter/10.1007%2F3-540-45547-7_3

Milroy, James & Leslie Milroy. 1999. Authority in Language: Investigating Language Prescription and Standardisation. London: Routledge.

Mishler, William & Richard Rose. 1997. “Trust, distrust and skepticism: Popular evaluations of civil and political institutions in post-communist societies”. Journal of Politics 59: 418–451.

Mudambi, Susan M. & David Schuff. 2010. “What makes a helpful online review? A study of customer reviews on Amazon.com.com”. MIS Quarterly 34(1): 185–200. http://wweb.uta.edu/management/Dr.Casper/Fall10/BSAD6314/BSAD%206314-Student%20Articles/Moderated%20Multiple%20Regression/moderation%20...pdf

Preece, Jenny, Blair Nonnecke & Dorine Andrews. 2004. “The top five reasons for lurking: Improving community experiences for everyone”. Computers in Human Behavior 20(2): 201–223.

Ren, Yuqing, Robert Kraut & Sara Kiesler. 2007. “Applying common identity and bond theory to design of online communities”. Organization Studies 28(3): 377–408.

Schlosser, Ann E. 2005. “Posting versus lurking: Communicating in a multiple audience context”. Journal of Consumer Research 32(2): 260–265.

Steinfield, Charles. 2004. “Explaining the underutilization of business-to-business e-commerce in geographically defined business clusters: The role of social capital”. Social Capital and Information Technology, ed. by Marleen Huysman & Volker Wulf, 209–230. Cambridge, MA and London: MIT Press.

Terveen, Loreen & Will Hill. 2002. “Beyond recommender systems: Helping people help each other”. Human-Computer Interaction in the New Millennium, ed. by J. M. Carroll, 487–509. New York: Addison-Wesley.

Wasko-McLure, Molly & Samer Faraj. 2000. “‘It is what one does’: Why people participate and help others in electronic communities of practice”. Journal of Strategic Information 9: 155–173.

Wasko-McLure, Molly & Samer Faraj. 2005. “Why should I share? Examining knowledge contribution in electronic networks of practice”. MIS Quarterly 29: 1–23.

Wenger, Etienne. 1998. Communities of Practice: Learning, Meaning, and Identity. Cambridge: Cambridge University Press.

Wenger, Etienne, Richard McDermott & William M. Snyder. 2002. Cultivating Communities of Practice. Boston: Harvard Business School Press.

Wiertz, Caroline & Ko de Ruyter. 2007. “Beyond the call of duty: Why customers contribute to firm-hosted commercial online communities”. Organization Studies 28: 347–376.

Zhou, Ming, Martin E. Dresner & Robert J. Windle. 2008. “Online reputation systems: Design and strategic practices”. Decision Support Systems 44(4): 785–797.