What do we Think an Encyclopaedia is?

The death of the encyclopaedia is increasingly reported in connection with the abandonment of hard copy reference publishing, the dispersal of library reference collections and the preference for end-users to seek information from search engines and social media. Yet this particular form of the book evolved in a very specific way to meet the needs of knowledge-seekers, needs which persist and perhaps flourish in an age of information curiosity. This article uncovers what is meant by ‘encyclopaedia’ by those who produce and use them. Based on survey and interview research carried out with publishers, librarians and higher education students, it demonstrates that certain physical features and qualities are associated with the encyclopaedia and continue to be valued by them. Having identified these qualities, the article then explores whether they apply to three incidences of electronic encyclopaedias, Britannica Online, The Stanford Encyclopedia of Philosophy and Wikipedia. Could it be that rather than falling into obsolescence, their valued qualities are being adopted by online forms of knowledge provision?


Why the Encyclopaedia?
When asked to picture an encyclopaedia, many will call up an image of a heavy, hardback book, to be consulted using its alphabetical headings, indexes and crossreferences to locate trusted pieces of knowledge.It is associated irrevocably with a mode of information-seeking which has been replaced by the use of the public web, or at the very least, powerfully-indexed online databases.According to Sundin & Haider (2013) 'We are now in the middle of a transition period and the way in which encyclopaedic knowledge, as a form of public knowledge, is communicated is changing profoundly'.The second decade of the 21 st century might, therefore, be an interesting time to address the question of how those most intimately associated with encyclopaedias express that it functions.This article 1 uses approaches drawn from the discipline of book history, to explore what the participants in the lifecycle of the contemporary encyclopaedia think an encyclopaedia should look like and the abstract qualities displayed by a good example.Publishers use the word encyclopaedia in their titles, librarians purchase them and consumers consult them with certain expectations of what they will find.The research on which this article is based consulted all three groups to see how they expressed what they thought an encyclopaedia was.
As outlined below, those interviewed were allowed to use their own words to describe the encyclopaedia and the approach of the research was exploratory and qualitative.Nevertheless, for the sake of clarity, an a priori definition will be useful to establish what kind of books this article is concerned with.For the purposes of this article, the encyclopaedia is a published reference book offering access to discursive factual information divided into entries and arranged systematically or alphabetically.It aims for a comprehensive coverage of a topic, or range of topics.Encyclopaedia contents are divided up into entries, but there is depth in the contents.They are more likely to be written in sentences than other forms of reference book and to be about the headword at the top of the entry.They might be considered the least referencey of reference books, containing the largest chunks of text and least reliance on page layout.This is contestable, indeed, is contradicted in places by the opinions of the participants in the research, but establishes a starting point for investigation.
Within the discipline of Book History, encyclopaedias have held a marginal place.A field that has at its centre questions about 'the reception, the composition, the material existence, and the cultural production of what is called the book' (Howsam 2006: 46) has tended to neglect those books designed for consultation rather than end-to-end reading.Far more interesting are celebrated works of fact or fiction, or, conversely, popular works with a domestic identity, objects of contemplation whose marks of ownership gives us clues to the cultural lives of their readers (Blair 2010: 230).Encyclopaedias have been the subject of study either as the means of exploring a specific historical cultural milieu, for example the French enlightenment (Darnton 1979) or early modern Europe (Yeo 2001), or as one of a series of methods of information storage whose identity as books is of small concern (Stockwell 2000;Wright 2007).
Yet encyclopaedias are an intriguing part of the life story of the book simply because of that which makes them distinct from other forms of published books.First, far more than fiction or monographs, they are identified by their format and physical appearance as much as by their content.They are immediately recognisable because of how they look: Instead of continuous pages of text, readers can expect to find individual articles or entries arranged under title headings.Prominent use of typographical features such as paragraphs and white space, bold and italic fonts and running heads mean that that they appear physically distinct from monographs.Secondly, encyclopaedia readers engage with them in a different way from books designed to be read from beginning to end.Unlike them, they are identified by their use rather than intention and a particular type of readerbehaviour, consultation, is associated with them (Attwooll 1986;Stevens 1987).Although scholars do use monographs and their indexes to refer to individual pieces of knowledge, encyclopaedias are specifically designed to meet this usage by facilitating access to their knowledge in the way they are arranged.Their function -by intention if not reception -is largely one of information rather than entertainment.This does not make them unique, but it means they lack a characteristic associated with such prominent forms of the book as the novel.
Encyclopaedias are also different from other reference books, such as dictionaries, gazetteers or recipe books.Encyclopaedias reveal far more about the society that produces them than forms such as the lexical dictionary, because their entries tend to be longer than those of dictionaries, use sentences, be discursive and to discuss rather than simply define their headword.Encyclopaedia entries are often articles covering many pages, aiming to encompass the breadth of an entire topic.This can make them controversial: Over the centuries, arguments about what an encyclopaedia should contain, who should compile it and how it should be arranged have reflected attitudes towards authorship and authority, the accessibility of knowledge and the possibility of capturing and recording all that is believed to be true and accurate.In some cases their influence has been huge: 'Grand projects like the Encyclopaedia Britannica, The Encyclopédie and The Oxford English Dictionary have all had tremendous social and cultural effects, acting as guardians of accuracy, setters of standards, summarizers of important and intellectual material' (Finkelstein & McCleery 2005: 4) (while this quotation includes a lexical dictionary, the OED is exceptional in terms of the amount of historical, one might almost suggest encyclopaedic, information it includes).It might be suggested that ubiquity of Google and Wikipedia mean they offer the same kind of influence as information resources today.In other contexts encyclopaedias are familiar domestic objects residing on the bookshelves of a family home.The authority conveyed by handsomely-bound volumes sitting on a domestic bookshelf, intended to offer access to factual information over the course of a lifetime, is also part of the encyclopaedia's story.
In either case, encyclopaedias have had an important role to play in the way people, and in particular, those who are not experts in any field, have chosen to acquire knowledge over the past three centuries.As Haider and Sundin (2010) suggest the 'encyclopaedic project' to share accepted public knowledge with a wider audience than the educational elite that gained currency in the early 20 th century, persists in the way that contemporary information-seekers seek to satisfy their curiosity, today using online sources including the open web.As the role of Book History is to explore the relationship between the book and its creators and readers, this article aims to investigate how those who participate in the encyclopaedia's production, communication and reception relate to it.
The research used in this article was largely based on user-responses to the hard copy encyclopaedia.This was because the intention of the research was to find a definition for a form of the book, which began its life in hard copy.The cultural notion of the encyclopaedia, when the term became first associated with the book form, was as an object, something printed, editorially mediated and held between covers.Nevertheless, the publishers, librarians and end-users interviewed are likely to have used electronic encyclopaedias, indeed may at this time exclusively use encyclopaedias online rather than in print.This does not negate the purpose of the research, which was to form a definition of what abstract and physical characteristics the encyclopaedia holds, according to its creators and users.But it is worth noting that the research may carry a disconnect between the encyclopaedias published, purchased and referred to on the one hand, and the popular idea of what one should be.This is reflective of a moment in Book History where certain types of book, of which the encyclopaedia is one, exist fully-formed as physical entities but are emerging in digital form as well.In many cases, they have ceased to exist in printed format.
The following sections outline the theoretical basis for the enquiry in the field of book history and the methodology used to carry out the research.

The Encyclopaedia and Book History
The focus of the field of Book History is the material form of the book.Donald F. McKenzie, a pioneer in Anglo-American book studies, redefined the field (then called bibliography) as studying 'texts as recorded forms, and the processes of their transmission, including their production and reception' (McKenzie 1985: 4).The work of the great book historians has helped to elucidate how the circumstances under which books have been produced, the intellectual context of their writing and their audience have influenced the form they have taken.An investigation into the material form and functional attributes of the encyclopaedia would [486] Culture Unbound, Volume 6, 2014 seem to be ideally placed within book history.Yet its focus has thus far mostly been individual titles, series or the works of specific authors.Unlike other forms of cultural analysis (such as Art History and Literary Criticism), the methodological tools of the discipline have rarely been used to establish the identity of a book genre or format.It is rare for a book historian to ask 'what type of book is an encyclopaedia?What form does it take?'.The research behind this article aims to redress this balance.Book History, a relatively new academic discipline drawn from history, bibliography, library and information studies, sociology and cultural and communications studies, is associated with no single methodology.Indeed, according to Finklestein and McLeery 'Competing methodologies are a feature of modern book history ' (2005: 12).Quantitative and qualitative analyses of sources such as print runs, employment records, bibliographies, libraries and booksellers' inventories was a common approach of the Histoire du Livre scholars of the 1980s and their followers (Darnton 1990: 162), revealing much about the selling, buying and reading habits of particular communities.As this approach came from the field of history, it has been most commonly used to investigate the publishing or reading habits of a specific place and time, rather than a type of book.
Analysing the text, or the object, has been another approach.McKenzie drew his research approaches from traditional analytical bibliography, studying books for the signs of textual intervention that were part of the book's transformation from an authorial text to the object that the reader encountered.His revolutionary approach was to assert that authorial intention 'must always be understood 'against a background of human conventions, expectations, practices and procedures ' (1985: 91).The text cannot be seen as emerging untainted from the author's individual genius.Rather, the author and printer combine to produce a text, which will fulfil the expectations of the consumer.Similarly, Gérard Genette's identification of paratexts, such mediating devices as title pages and book jackets, as 'zones of transaction' between the author (or publisher) and the reader (Genette 1997: 2) has been an influential way of considering the book as a physical object.Such features are the means by which a book is packaged to convey its content in a particular way to a reader.However, Genette chooses to use his method on celebrated works of French literature and the approach has not been applied to identify the zones of transaction across a category of book.
Sociological and ethnographic research is also part of Book History, particularly answering questions about reading habits.Reader-response criticism was adopted from cultural theory by scholars such as Janice Radway, who interviewed the readers of romance novels as a means of considering this form of publication (Radway 1984).A range of qualitative and quantitative approaches using surveys and interviews have answered questions about the role that certain books play in the lives of their readers.Here, as was the case with Radway's research, the focus has sometimes been specific genres, but this is usually within fiction.
The research for this article aimed to gather opinion on the nature of the encyclopaedia from those involved in producing, communicating and consuming it, providing an illustration of what an encyclopaedia is considered to be by those who are most familiar with it.To some extent, it followed Radway in identifying a group of users of a particular type of genre fiction and recording their reactions.As with her research, the present author aimed to examine how sets of conventions associated with a particular type of book were perceived by its users.However, the key elements of narrative, plot and characters are not part of the encyclopaedia, meaning that a genre studies approach would not have been appropriate.Nor do users of reference works have the same kind of emotional relationship with them as those of forms of fiction.As Blair (2010: 230) points out, encyclopaedias are often owned by institutions rather than by individuals, providing few clues to their place in their readers' lives (although there is of course, a long history of encyclopaedias sold to families, often paid for by instalments -see Einbinder (1964) for a critical account of this).In many cases, encyclopaedias are purchased by librarians, who use them for research themselves, as well as recommending their use to readers.It seemed unlikely that many users, could a suitable sample be found, would be able to produce a detailed articulate response on the encyclopaedia, as Radway's reader-group had been able to.An ethnographic approach examining encyclopaedia users' response to the form was therefore rejected in favour of using surveys and brief interviews with a small number of questions, albeit ones encouraging free expression and without pre-determined choices.Moreover, it was felt that in the world of the encyclopaedia, the end-user's opinion alone was insufficient.This study therefore sought representatives from all parts of the encyclopaedia's life-cycle.
In order to identify the key participants in the life of the encyclopaedia, Robert Darnton's 'Communication circuit of the book' outlined in his article 'What is the history of books' (Darnton 1982), was used as a framework.His circuit identifies the people or industry functions, which contribute to the book production and consumption process including authors, publishers, printers and readers.He depicts external factors ('intellectual influences' 'economic and social conjuncture' and 'political and legal sanctions') affecting all part of the cycle.Although some areas are kept broad ('Readers' includes 'purchasers', 'borrowers', 'clubs' and 'libraries') the diagram is necessarily based on Darnton's own time and place of interest, that is, the French enlightenment.Darnton's circuit produced a critical response from Thomas Adams and Nicolas Barker (1993) who countered with their 'bio-bibliographic model'.Where Darnton highlighted the roles of a book's producers and consumers, Adams and Barker pinpointed stages of the life of the book itself: publication, manufacture, distribution, reception and survival.As such, their circuit is more universal, but in some ways less descriptive.The model the author produced for the contemporary encyclopaedia (Figure 1) brings in elements from both approaches, highlighting both actors and processes.This was [488] Culture Unbound, Volume 6, 2014 used as a basis for identifying the key participants in the encyclopaedia's lifecycle, and therefore, the sample for interviews and surveys.

Data Collection
The data for this article was collected from the three main participant groups in the encyclopaedia's communications circuit: publishers, librarians and readers (perhaps more accurately called users, given the manner in which encyclopaedia contents are commonly accessed).A mixture of interviewing and surveys was carried out for a number of reasons, some of them practical.While it was possible to interview the publishers group, a survey was the only method to reach the much-larger number in the reader group in the time available.In all three cases, subjects were asked generally about their experience in connection with reference books, then specifically encyclopaedias, in order to focus their minds on the subject of the survey.They were then asked to describe in their own words what they considered, in turn, the physical features and functional values (or attributes), they would expect from an encyclopaedia.
Although initial questions differed between the three groups, because of the differing background knowledge on encyclopaedias they held, all three were Culture Unbound, Volume 6, 2014 [489] asked the same three open questions, whether interviewed in person, by email or by online survey, as follows: • What physical characteristics make up an encyclopaedia to you?
• Can you think of any physical features which *must* be present for it to be an encyclopaedia?
• Can you tell me what abstract qualities characterise a successful example of an encyclopaedia?
Opening questions for the publishers and librarians were about their experiences with encyclopaedias, how they currently or previously had worked with them in their professional lives.Publisher interviews were carried out in person, by telephone or by email and all subjects were asked to read and amend the record of the interview to ensure it was representative of their opinions.Where a questionnaire was emailed, follow-up questions were used and the participants were able to check their answers, as with the interviews that took place in person.The aim was to gather an accurate impression of their beliefs about encyclopaedias, even if they changed their mind between the original interview and subsequent reading, and even if this actually gathered what they felt they ought to have said, rather than their first impressions.Like the publishers, the librarians who responded all had a high degree of awareness of the role of the encyclopaedia and were thus able to describe clearly their expectations.The questions put to them regarding their opinions of encyclopaedias (but not their experience) were identical to those put to publishers.Most completed questionnaires by email, but two were interviewed in person.Again, they were all allowed to correct the record of the interview.
The encyclopaedia readers were surveyed online.Initial questions were introduced, not to gather data, but to prepare the subject for the questions that followed.For example, they were initially asked about the types of reference books that they used or owned.This data was not used, but was aimed to help them identify in their minds what they understood a reference work to be.They were also asked what they considered to be the distinction between a dictionary and an encyclopaedia.This gained interesting results, which were not analysed, but aimed to help the subjects consider what makes reference books distinct from each other.
Every statement of opinion from each interview record or survey result was extracted and tabulated.Statements that seemed to have similar meanings were grouped together and labelled.Figure 2, below illustrates one example of how the free-text statements from across the three sets of participants were grouped under a single heading inductively selected by the author.This type of content analysis is inherently problematic.It is impossible to ascertain whether one individual's answer describes the same thing as another's and both are filtered through the analysts' subjectivity.Even where subjects used identical vocabulary, there was no guarantee that two people meant the same thing when they use the same word. [490] Culture Unbound, Volume 6, 2014 Moreover, while the users were all based in English-speaking countries, English may not have been the first language of some of them.In fact, quasi-synonymous concepts were grouped together because the underlying encyclopaedic characteristic was interpreted as being essentially the same.The results are a compromise between the ethnographic approach, where every result is considered uniquely valuable, and the universalist, which aims to create a consensus across the answers.Each participant's answer has an ethnographic value, both for the individual and the group they represented, but commonalities were sought between the answers to provide an overall picture of the encyclopaedia according to its communications circuit participants.

Research Sample
Finding a representative sample of interviewees who can be relied upon to provide honest answers to questions asked is a key challenge and was the motivation for identifying the encyclopaedia communications circuit as a guide.The stages of the communications circuit merge and blend into one another and participants take up multiple roles, or switch between them.However, for the sake of clarity, three types of participants were identified: publishers (which includes writers and researchers who participate in the creation of the book), librarians and users, all of whom needed to have an intimate relationship with encyclopaedias as part of their roles.Publishers, which included commissioning, consultant and contributing editors, were found through word of mouth and by making direct approaches.This group represents a small community and this was the best method to track down those with direct experience of encyclopaedias.Librarians were approached using email discussion lists and the online social network Twitter.This gave potential access to a large number of potential professionals although a comparatively small number agreed to complete the survey.No printers, developers or booksellers who identified themselves in relation to the encyclopaedia could be found.
The selection limited the possible number in each groups likely to be able to provide responses.The numbers found for each group varied: 12 responses (out of 20 approached) in the publisher category', 13 (out of 24 who began the survey) librarians and 85 users.The comparatively small number of publishing industry professionals, drawn from both UK and US publishing companies, was nevertheless a large proportion of those working within the reference publishing industries of those countries as a whole.The much larger number of end-users interviewed indicated the larger community from which they came.The librarians, all working in the UK, were a small but vocal sample of the community who engage with encyclopaedias as part of their work.More would have been preferable, but were unobtainable in the time available.This was mitigated by the fact that, in the cases of both the librarians and publishers, a good spread across different functions (editing, marketing, acquisition, research) was achieved.The publishers were in Culture Unbound, Volume 6, 2014 [491] many cases freelance, but those who were employed had been so in six different publishing companies.The librarians were all employed in different institutions.Encyclopaedia readers presented a challenge for the sample: to find respondents who considered themselves encyclopaedia users and were capable of articulating their thoughts about them.While the survey would ideally have had responses from users from a range of backgrounds, including non-academic users, it was anticipated that the higher education community would contain a higher number of encyclopaedia users who identified themselves as such.In addition, it would have been extremely difficult to ensure that any selection of 'ordinary members of the public' was random within the scale of the study.Instead, the reader sample was taken from a distinct group, postgraduate Students from two UK multi-disciplinary universities, contacted via their postgraduate school / deanery.It is recognised that they could only be representative of their own grouping, not of the user category as a whole, but postgraduate students are more likely than undergraduates or those not in higher education to have used reference works and formed an understanding of what an encyclopaedia is.To an extent, all three sets of answers represented the opinion of an 'elite', in this case, those most identified with the encyclopaedia as creators, communicators and users.

Results
Respondents were asked both what they thought the most important abstract quality or value in an encyclopaedia was (here called 'function', to distinguish it from a physical characteristic) and what physical features they would expect to find in them.The three groups spontaneously named many of the same things, although the language varied within and across the groups.The chart below (Figure 2) illustrates this by showing the spontaneous answers individuals gave, subsequently categorised under the heading 'Authority'.Clearly this categorisation is challengeable, but indicates that even described differently, the different groups shared some of the same concerns: [492] Culture Unbound, Volume 6, 2014 This figure illustrates the process taken to categorise the statements made by the respondents and interviewees.Some of the distinctions between the way the function is described can be ascribed to the different ways they encountered the question: a face-to-face or telephone interview is likelier to engender a wordier response than an online survey which the anonymous participant may like to complete as quickly as possible.Under those circumstances, it is interesting to note how far the three groups were in agreement both in their choice of function and the way they chose to describe it.Figure 3 was produced by categorising and counting the statements participants made in answer to the questions about an encyclopaedia's functional attributes, then placing each attribute in order of number of mentions.A 'mention' was identified as any descriptive term noted by any respondent in answer to the questions asked, so that if their answers included more than one functional attribute, these were counted separately.See Figure 2, above, for how descriptive terms and statements were tabulated before being counted.
A number of functions were valued prominently by all three groups: Authority, accuracy, ease of reading, structure / accessibility and comprehensivity were the most common.This seems consistent with other sources, for example, guides designed to help librarians select reference books for their collection.Louis Shores' influential publication Basic Reference Books (Shores 1939) or the contemporary Cassell and Hiremath's Reference and Information Services in the 21st Century (Cassell and Hiremath 2009), most commonly advise librarians to evaluate each title for authority, accessibility, clarity of purpose, good physical format, currency, style, originality, suitability for audience, accuracy and bias. 2 The same functions, authority, accessibility, accuracy and quality of publication recur to define [494] Culture Unbound, Volume 6, 2014 what an encyclopaedia should be valued for.This indicates that the notion of what a good encyclopaedia embodies is culturally shared among encyclopaedia users and creators.Yet there are interesting differences both within and across the three groups.Readers were more likely than the other two groups to include as functions things which might be considered to be physical features, for example, 'crossreferencing' and 'indexing', an indication that the distinction between physical and abstract features is less important to them.They were also the only group to identify 'brevity' or 'conciseness' as a desirable attribute, perhaps reflecting a concern to conserve the time they spend seeking information, something not always recognised by the publishers who produce their reading material.It is notable that another item mentioned only twice by end-users and not at all by other users is 'neutrality' or 'lack of bias'.Given the high position of 'authority' and 'accuracy', the ability to trust the work is clearly a concern for all three groups, but it did not occur to any publishers or librarians to specify objectivity as desirable.Users were also alone in identifying 'range' as desirable function, using such phrases as 'Unlimited topics', 'All-encompassing knowledge', 'Varied information' and 'Broad range of coverage'.It may be that those with a professional identification with encyclopaedias, the publishers and librarians, took this quality (and others) for granted.Range, the notion that an encyclopaedia should embrace a variety of topics, or a single topic with a breadth of scope, is for some the very definition of the term 'encyclopaedia'.
The librarians also placed more emphasis on certain aspects of an encyclopaedia than other groups in their answers.A number of respondents were concerned with how the encyclopaedia matched the needs of its audience.Responses included phrases such as 'Content should be pitched at different levels i.e. brief and simple overview for beginners and longer, more in depth articles for researchers' and 'Foremost, I think an encyclopedia needs to be informative to the level it is aimed at, relevant'.These indicate the role librarians take as the intermediaries between the information source (the encyclopaedia) and the user (which, for an academic library, would be students).It was, however, also mentioned by two of the publishers, using the phrases 'Usefulness to the user' and 'The most effective reference publications are those that understand the needs of their target audience'.Of course it is a concern of the user too, but was perhaps reflected in some of the other responses, such as 'ease of use', or the equal number of users who responded that they expected 'brevity' or 'succinctness' as well as 'depth of information' or 'detailed information'.In all cases, the suggestion is that the encyclopaedia should cover its topic in a way that is just right for its audience.This is perhaps the defining feature of the encyclopaedia, sitting between the monograph, which may have too much detail, and the dictionary, which may not have enough.
When it came to the physical features expected in an encyclopaedia, there was a far wider range of answers within and across the three groups and a long tail of characteristics mentioned only once or twice.Figure 4, below, shows all the items mentioned across the three groups.4 was produced by categorising and counting the statements participants made about an encyclopaedia's physical features, then placing them in order of most mentions.As with Figure 3, each descriptive statement or term was counted separately, even where respondents gave more than one answer to the question.
Across the three groups, the items most expected in an encyclopaedia are an index, arrangement into entries, citations, cross-references, contents page and alphabetical order.Also scoring highly is that it should be a hardback book of considerable size and weight.The higher-scoring features largely fall into two groups: those which help the user locate the information within the book, and those concerned with the encyclopaedia as a physical object.The fact that participants from all three groups mention features like index, entry arrangement, structured organisation and alphabetical order, indicates the strong identification of an encyclopaedia as a non-sequentially-accessed book, dependent on its structure to be useful.Meanwhile, the high score of hardback binding and large weight and size show that it is expected to be a substantial object.The physicality of the object is emphasised by some of the features that appear in the long tail such as cloth placemarker, thumbmarks and good quality or (conversely) thin paper.For these partic-ipants, the encyclopaedia has a distinct physical identity separate from its informational content.
Some of the differences between the groups' choice of features relate to their articulacy on the subject: publishing professionals have a clearer notion of the parts which comprise a book and a more technical vocabulary than users, although interestingly, in places they identified the same feature in different words.For example, what one publisher described as 'an 'onion skin' approach to article layout, where a summary precedes more detailed sections, is very similar to what a reader called 'synopses for each section'.Even the differences between some of the tail of responses in fact show a similar approach: for example, the users' 'anything to make things clear' (summarised above as 'clarifying features') is an indication of wanting to find what they are looking for without difficulty.This relates strongly to the publishers' concerns around good structure and access, described in such ways as 'Digestible access', 'Organisation enabling information to be accessed non-sequentially' and 'Systematic organisation'.The 'lots of text', mentioned by one librarian and one user is consistent with the more-frequently mentioned 'compendiousness' and 'in-depth coverage'.Allowing for interpretation of the different language, there is a surprising consistency of opinion as to what physical form an encyclopaedia should take.
All three groups considered it important as to whether an encyclopaedia should be subject specific or cover a range of topics, that is, whether it should be, for example, 'an encyclopaedia of children's literature' or a general encyclopaedia attempting to cover all subjects.Both features are associated with encyclopaedias, but subject specificity was mentioned more often, by both users and librarians.Similarly, some of the answers regarding the length of an entry show a divergence of opinion.Reflecting the contradictory responses whereby end-users wished for the functional attributes of 'brevity' as well as 'detailed information', five endusers mentioned 'brief' or 'summarised' entries as features they would expect and a number of librarians and publishers expected 'in depth' (or 'essay style') entries.Such contradictions have historically concerned writers about encyclopaedias.In the same edition of American Behavioral Scientist, encyclopaedist Charles Van Doren described encyclopaedias as having a 'a tradition of dedication to truth and completeness' (Van Doren 1962), while cultural historian Jacques Barzun warned encyclopaedia compilers only to cover the ideas which 'have engaged the protracted attention of mankind' (Barzun 1962).In a sense this paradox embodies the central contradiction at the heart of the encyclopaedia: to be comprehensive, but only to select what is useful and trustworthy.
Even allowing for the differences in languages and priorities, the results indicate a shared understanding of what an encyclopaedia is among those who create, purchase and consult it.Among the three groups, there was substantial agreement that a good encyclopaedia might be expected to hold the following characteristics: • be a substantial physical object Culture Unbound, Volume 6, 2014 [497] • contain information organised into entries • contain articles in alphabetical order • organised in such a way that items of information are easy to find • treat topics in some depth • be accurate • be authoritative • be well-written • be comprehensive • presented in a manner appropriate for its audience.
Although many of these features might be found in other types of book, aggregated they represent a very distinct type of book, recognisable across the contemporary encyclopaedia communications circuit.This is intriguing, particularly when certain physical features may seem old-fashioned when most publishers' encyclopaedia profits are now made online (Bookseller 2011).All three groups have a clear idea what an encyclopaedia should be like and how it should function, and there is a certain amount of consensus across all three groups.Their vision of the encyclopaedia is not a revolutionary one, but it demonstrates that a culturally shared notion of the encyclopaedia exists among its creators, communicators and users.At a time when, as Sundin and Haider (2013) put it, 'the use of encyclopaedic knowledge has become different, always available and in constant competition with other sources', in this study, all parts of the encyclopaedia's communications circuit still identify it as a distinct form of the book.

Encyclopaedias in the Digital Age
Yet, the existence of the encyclopaedia online cannot be ignored.Without prompting, some respondents suggested features only to be found in an online encyclopaedia, such as 'a decent search feature' and 'graphic enhancements' (publishers) and 'Attractive online display without too many flashing distractions' (librarians), not to mention the two end-users who responded that an encyclopaedia should be online (one using words not repeatable in this article).As mentioned previously, the participants are perhaps more likely to have used online encyclopaedias than hard copy even while they identify physical features, such as hardback binding and cloth placemarkers which would only be found in a physical object.
Evidence suggests that opportunities to use hard copy encyclopaedias are becoming rarer.Libraries have increasingly directed their scarce resources away from hard-copy encyclopaedias, towards online products, searchable and available 24-hours a day without the need for staff (Bradford 2005;Heintzelman, Moore & Ward 2008;O'Gorman & Trott 2009).Publishers increasingly release their titles [498] Culture Unbound, Volume 6, 2014 in online form only (Bond 2008;Jones 2008) and the decline in print sales in the late 2000s was noted by one publisher as 'dizzying' (Danford 2009).Social media presents challenges too.When the answers to questions can be crowdsourced through blogs or social media sites such as Twitter and Facebook, the question arises as to why anyone would need reference source mediated by the publication process.Featherstone and Venn (2006) suggest that social media sees itself as offering an opportunity for 'de-authorization of the cult of experts', in particular Wikipedia, which since 2001 has offered encyclopaedia-style articles, written by voluntary contributors and editable by almost anyone, for free via the web (Wikipedia 2013a).
According to Haider and Sundin (2010) 'One could be tempted to think that the encyclopaedic notion would go out of fashion when (Web) search engines create instant access to most digital content.'Yet the desire to find trustworthy information on a topic gathered together under a specific article heading has persisted and continues to be catered for.A quick glance at three examples, Britannica Online which remains, more or less, an editorially mediated publication online, The Stanford Encyclopedia of Philosophy (SEP), which is looser in publication structure, has no hard copy roots, but is written by experts and Wikipedia, show that they imitate hard copy encyclopaedias: they divide their contents into articles, use both bold and sub-headings, employ summary sections or synopses, include bibliographies and citations, offer alphabetical access, indexes and crossreferences.This is consistent with Bolter and Grusin's concept of 'remediation' (2000) whereby new media formats ape older formats in the course of refashioning them.Of course online encyclopaedias do not have hard copy bindings or cloth placemarkers, but the pages still look like encyclopaedia pages.Many of these features might be considered redundant when the contents are accessible through a free-text search, but readers still want to find information under discrete article headings, to find further sources of reading, via citations and references, and to draw relationships between the topics they cover using cross-references.Moreover, all three describe themselves as 'encyclop(a)edias', suggesting that they desire the associations connected with the term.
Abstract notions of accuracy and authority also persist into digital forms.The background material on each site (particularly copious on Wikipedia) promotes the contents of each online encyclopaedia.Britannica, for example, announces 'In a world where questionable information is rampant, we provide products that inspire confidence, with content people can trust' (Britannica 2013), illustrating the point that 'Old trustworthiness -tied up in tradition, expertise and local relevance -gains new currency in networked settings' (Sundin & Haider 2013).SEP says 'From its inception, the SEP was designed so that each entry is maintained and kept up to date by an expert or group of experts in the field.' (Stanford Encyclopedia of Philosophy 2013).While Wikipedia emphasises that 'People of all ages, cultures and backgrounds can add or edit article prose, references, images and other media' they also add that content needs to be 'verifiable against a published reliable source' and that older articles, which have seen additions from 'experienced editors' are more reliable than newer ones (Wikipedia 2013a).There is an immense amount of background material on Wikipedia explaining and justifying their publishing model as a good method of producing accurate encyclopaedia entries (far more so than in Britannica, where the publication's 300-year history and well-known name might be assumed to speak for itself).Their argument is not that accuracy and authority are not important in an encyclopaedia, but that they can be provided by an alternative publishing model.Even the much-quoted Nature Magazine comparison between Wikipedia and Encyclopaedia Britannica (Giles 2006) was essentially a competition to prove the accuracy of one source over the other.
There are valued notions associated with the encyclopaedia that Wikipedia does not embrace.Unlike Britannica and SEP, there is no editorial masterplan dictating how it should be structured and the scope it should embrace.This has attracted criticism, suggesting that its coverage is skewed towards certain types of topic 'where there is a wide distributed knowledge base and a large a pool of people with time on their hands to contribute' (Publishers Weekly 2009).One of Wikipedia's own boasts is that it embraces a wider range of topics than those constrained by editorial or academic needs, and there is no limit to its potential length or the number of topics it can embrace.Indeed, the 'What Wikipedia is not' page states 'there is no practical limit to the number of topics Wikipedia can cover' (Wikipedia 2013b).The process of continuous, collaborative revision might, over time, mean that the encyclopaedia as a whole, or any topic or single article could grow infinitely over time as more people contribute.This contradicts the possibly defining function of the encyclopaedia, to cover a subject in depth, but with concision and at the appropriate level for a specific audience.Yet there is more editorial intervention in Wikipedia's structure than might be thought.A Quality Assessment team assigns a classification tag to each article which might suggest it is, for example a 'stub' or a 'featured article' (Wikipedia 2013c) while the 'Categorization' guidelines, among many other suggestions of how volunteers should approach writing an article, (Wikipedia 2013d) are copious.

Concluding Remarks
This article illustrates the features expected in an encyclopaedia by those who produce, communicate and use them at a time when this form of the book is increasingly coming to exist in online form only.The participants in the research identify qualities and attributes which can be seen or are boasted of in prominent digital encyclopaedias, suggesting that the online information offering draws much from the hard copy world which preceded it. [500] Culture Unbound, Volume 6, 2014 While it appears, from these observations, that publishers aim to meet user expectations of an online encyclopaedia by echoing or emphasising some of the physical and abstract features of the hard copy form of the book, there is further research to be done in this area.In a world where highly sophisticated approaches to web design and usability exist, it would be interesting to explore the decisions made by publishers in creating online encyclopaedia environments.Does their design reflect in any way the iconic physical item?Similarly, while there is very little research into user-behaviour in relation to hard copy encyclopaedias (Bradford 2005 is a rare exception), user-studies could be carried out, or metrics examined, to explore how their online equivalents are used.And it would be interesting to explore whether design and use of encyclopaedias differed from any other online reference sources.These questions, however, were beyond the scope of the research used in this article.
The generic features of the encyclopaedia remain recognisable to those most clearly connected to its production and consumption.The clarity of its identity seems only to be confirmed by the fact that many remain in digital forms.Exploring how far the characteristics of different types of books persist and alter in the online world has much to tell us about publishing, communication and reading.The suggestion that book types can be more than objects, can transcend the physical and persist into the digital world adds an intriguing frisson to our understanding of book forms' relationships to their users.It suggests that users continue to bring expectations developed in the physical world to the way they interact with their digital proxies.For many interviewed as part of the research for this article, the concepts of creating, organising and finding encyclopaedic knowledge developed with the hard copy book.However, digital natives were also among those interviewed, and they still recognise the encyclopaedia as a distinct book form.How this might develop in the future remains to be seen and there is no doubt that information seeking and provision is in a state of flux.However, the appetite for authoritative and accurate content organised into easy-to-navigate articles appears not to have diminished, and it continues to be provided for online.
Katharine Schopflin recently successfully defended her PhD thesis at University College London's Department of Information Studies.Her topic was the nature of the encyclopaedia as a form of the book, exploring both the qualities associated with a successful example of the encyclopaedia and the physical features characteristic and unique to it.She is additionally a library professional practitioner, currently working as a knowledge manager for a membership organisation in London.She has published numerous articles in the information press and is the author of A Handbook for Media Librarians and the upcoming A Handbook for Corporate Information Professionals both published by Facet Publishing.E-mail: katharine.schopflin@gmail.com.
Culture Unbound, Volume 6, 2014 [501] 1 The research for this article was originally carried out as part of a PhD project which considered a broader spectra of questions.
2 An analysis of 11 such guides appeared in the PhD thesis from which the other research in this chapter was taken.

Figure 1 :
Figure 1: The Encyclopaedia Communications Circuit

Figure 2 :
Figure 2: Example of answers given grouped together under the category 'Authority'

Figure 3 :
Figure 3: Functional attributes mentioned by participants, shown by numbers of mention and by percentage

Figure 4 :
Figure 4: the most common physical features expected in an encyclopaedia

Figure
Figure4was produced by categorising and counting the statements participants made about an encyclopaedia's physical features, then placing them in order of most mentions.As with Figure3, each descriptive statement or term was counted separately, even where respondents gave more than one answer to the question.Across the three groups, the items most expected in an encyclopaedia are an index, arrangement into entries, citations, cross-references, contents page and alphabetical order.Also scoring highly is that it should be a hardback book of considerable size and weight.The higher-scoring features largely fall into two groups: those which help the user locate the information within the book, and those concerned with the encyclopaedia as a physical object.The fact that participants from all three groups mention features like index, entry arrangement, structured organisation and alphabetical order, indicates the strong identification of an encyclopaedia as a non-sequentially-accessed book, dependent on its structure to be useful.Meanwhile, the high score of hardback binding and large weight and size show that it is expected to be a substantial object.The physicality of the object is emphasised by some of the features that appear in the long tail such as cloth placemarker, thumbmarks and good quality or (conversely) thin paper.For these partic-