Introduction to the Second Edition
Introduction to the Second Edition
The best way to understand this introduction to the second edition of International Encyclopedia of Linguistics (IEL) is to read William Bright's introduction to the first edition. The approach to and structure of this new edition are much the same as those of the previous one. My goal was to cover the entire discipline through cogent articles written by experts in language accessible to everyone, with the entries organized to reflect the conceptual dependencies of the field, thoroughly cross-referenced, easily searched and user friendly, while vigilantly preserving theoretical balance and neutrality.
Knowing that in linguistics, no error or bad choice, however minor, goes unreviewed, I nonetheless signed on to the same impossible task as Bright. But my job was much easier because I had Bright's extensive, careful work as a base. Those of us who have toiled in the reference industry know how essential a solid first edition is to future versions of a reference work. I know that I speak for the whole field in acknowledging Bill Bright's important role as originator of International Encyclopedia of Linguistics and, more generally, as patient overseer of the discipline.
Because of the substantial overlap between editions in organization and motivation, this introduction can be most useful by charting the divergences between the organization of these new volumes and Bright's. How does this edition compare with the first?
The second edition of IEL contains 957 articles and about half as many more headwords and subheadwords, given an increase in blind entries (place-holder entries that point to others for their content). All the articles from the first edition are updated in some way, with most ranging from 40 percent revision to complete rewriting. About 15 percent of the articles are new and about 3 percent of the first-edition entries were eliminated from the second. The extensive updating of a ten-year-old reference work slated to endure at least another decade led to the addition of some theories and subfields, the elimination of others, expanded treatment of some existing concepts because of their rising importance, and reduced treatment of others because of their increasingly limited role in the field. Decisions about these matters were made with every attempt to keep to the high ground and to follow two reasonable dictates: first, not only include the core of the field and ideas of current influence, but also try to anticipate what will be current in the years to come; second, try to avoid pet theories and personal exposés. The first edition was especially strong on languages of the world and social and anthropological approaches to language. The second edition has maintained these strengths and given attention to new directions in the field, such as Optimality Theory, the Minimalist Program, the widening influence of functional and typological linguistics, the spreading impact of discourse analysis on subfields outside discourse proper (applied linguistics, the computational linguistics, semantics), increasing detail in findings in formal linguistics, marked changes in applied linguistics (perhaps the most extensively revised topic area), and advances in computational and mathematical linguistics.
In the decade since the first edition appeared, linguistics has moved closer and closer to psychological and neurobiological inquiry, and so this edition has coverage of the various cognitive and evolutionary approaches to language, including neurolinguistics and brain imaging, cognitive science, critical periods of acquisition, linguistic relativity, learnability, and language disorders. Many subfields and concepts have been “unpacked,” with their constituent ideas given explicit treatment: for example, there is a separate composite entry on phonological processes, with full treatment of assimilation, dissimilation, and so on. There are also major additions to the coverage of languages, with new articles on Zulu, Wolof, Khoisan, American Sign Language, world Englishes, and artificial languages. In the end, the guiding idea for coverage in IEL was to maximize information (“less is bore!”) while preserving readability.
Entries and organization
Choices about what entries to include and the structure of the Encyclopedia were guided by three main principles:
1. Choose entries that are neither too general nor too specific. Which of the following terms should be a headword in the Encyclopedia: sound, phoneme, or delayed release? The natural response is that IEL should include all of them, but if it does, the volumes immediately begin to grow in length exponentially, thus sacrificing usability for coverage. The second edition tries to reach a middle ground between concepts that are highly abstract and those that are narrow and often theory specific, no matter how important any of them might be: phoneme is included, but sound is too general and delayed release too specific. In most cases, narrow concepts are embedded in articles and so can be found through the index. But the user will not find entries for language, structure, speech, or linguistics; “Grammar” and “Meaning” are blind entries, that is, placeholders that point to other entries for exposition. There are no separate entries for INFL, regime, or garden path, for example, influential concepts to be sure, but concepts that can be found under entries in the subfields in which these concepts occur (in these cases, “Phrase Structure,” “Optimality Theory,” and “Parsing”).
2. Organize entries to reflect the conceptual structure of the field. On the whole, the second edition is more complex in organization than the first edition. Much effort was put into renaming entries (“Natural Language Processing” has become “Computational Linguistics”), inserting new entries for balance (“Morpheme” requires “Phoneme”), and reorganizing entries to reflect the way the concepts in the field relate to and depend on one another (“Neurolinguistics” has changed from a blind entry to a composite entry; “Discourse Markers” is a separate entry because it is an influential concept in its own right; “Applied Linguistics” is significantly less complex than might be expected because its potential subentries are all independent notions). There is significantly more nesting (entry with subentry and subsubentry). For example, “Language Change” is a new composite entry subsuming all the approaches thereto; “Pragmatics and Contextual Semantics” has been split, and split again, to give explicit treatment of implicature, pre-supposition, and related notions; “Semantics,” “Philosophy of Language,” “Institutional Linguistics,” and “Linguistics and Literature” are highly complex entries, subsuming various approaches and concepts given independent treatment in the first edition. Overall, the goal with this “complexification” of the Encyclopedia was to promote one-stop shopping, so that a user might find not only what he or she is looking for at an entry, but more of what the field itself sees as related to the notion under search.
3. Ensure that entry choice and organization promote ease of use. Nothing makes a user of a reference work more frustrated than to have a term or concept in mind, but never to be able to locate that term in the reference work— or to have to struggle so much to locate the term or concept that it might be more advisable to abandon the search altogether rather than come away from the search successful, but dazed. How to avoid these pitfalls? The second edition applies a number of heuristics to meet this challenge. The first edition limited entries at L and S, concerned with what lexicographers affectionately think of as the clumping problem, clusters of entries at particular headwords. An encyclopedia of linguistics would seem to have clumping at L and S: language, linguistics, speech, and so on. The second edition allowed L- and S- entries, and without much consequence. There are many more C-, P-, and S- entries anyway, whatever the policy for L and S. What this means is that entries in the first edition that had been modified to avoid L or S now appear at those letters: “Language Attitudes,” rather than “Attitudes to Language,” “Speech Perception” rather than “Perception of Speech.” “Language Acquisition” is a blind entry, pointing to “Acquisition of Language.” This was a deliberate choice because it was thought that users would have both language and acquisition in mind when searching for information on language development.
One peculiarly technical issue arose in considerations of headword choice. There are many terms in linguistics that are ambiguous across subfields: free and bound are technical concepts in both morphology and the syntax and semantics of anaphora; case is a term in both formal syntax and semantics; feature has different meanings in phonology, syntax, and semantics; local, locality, and localization mean certain things in semantics or grammatical theory, and quite other things in neurolinguistics; declarative has one meaning in syntax and a different one in computational linguistics. Every attempt was made in the second edition to make these ambiguities explicit and to point the user to the appropriate entries for these different meanings.
Finally, the entry list includes a number of new symbols and abbreviations: “AAVE,” “OT,” “LF,” “GPSG,” and “TAG.” Notation has a significant place in linguistics, and a user of the Encyclopedia would no doubt have occasion to search the volumes with these nonlexical and nonphrasal entries in mind.
There are many more blind entries in the second edition than in the first in order to capture what was thought to be the search procedures of an intelligent user. But this again raises a selection issue: how to decide which terms should be blind entries? The second edition contains three major kinds of blind entries:
1. Truly empty terms whose content is found elsewhere. An example of this kind of blind entry is “AAVE,” which directs the reader to a full entry for “African American Vernacular English.” Other examples are “Assessment” pointing to “Language Testing” and “Onomatopoeia” pointing to “Sound Symbolism.”
2. Mid-level concepts that evoke higher-order terms (their full and partial superordinates) but do not merit separate treatment as discrete entries. These are by far the largest class and the most complicated to justify. An example is “Agent,” a core concept within semantics and grammatical theory and one that is included as a blind entry pointing to “Case” and “Thematic Roles.” The editorial concern here was whether such terms— for example, isogloss, perlocutionary act, and usage— ought to be free standing. In the end this was a judgment call and the rule of thumb was to include as blind entries concepts that might be independently searched for but that do not stand entirely on their own conceptually and that make better sense for exposition in the entries on their superordinates. Thus, “Tiers” is a blind entry because it has semi-independent status in the field but is understood principally through “Autosegmental Phonology.” The term tiers contrasts with a term like tableau, which the IEL user is likely to search for already knowing that it is part of “Optimality Theory” and so intrinsically bound to its superordinate concept.
3. Concepts that, for epistemological or socio-political reasons intrinsic to the field at the moment, are too complicated to include as separate entries, despite their importance and status as a likely target of independent search. A classical example of this type is “Syntax,” an entry that gives no particular definition but points the reader to subfields and theories, where fuller, independent treatments can be found. Other examples are “Clinical Linguistics,” “Competence,” “Modularity,” and “Representation.”
Blind entries are signaled by See, and perhaps the best way to appreciate the three major classes of them is to think about what See means for each of these types. Category 1 construes See as equals: “LF” equals “Logical Form.” Category 2 construes See as is a part of or is best located under the larger concept of: “Perlocutionary Act” is best located under the larger concept of “Pragmatics and Contextual Semantics.” Category 3 construes See as is best understood in the convergence of the concepts aggregated under: “Transformational Grammar” is best understood in the convergence of the concepts aggregated under “Formal Grammar,” “Principles and Parameters,” “Minimalist Program,” and “Transformations.”
Blind entries were one area of editorial decision that could not be settled beforehand, but had to be managed as the volumes unfolded. Many of these choices depended on what authors were saying about terms and concepts and could not be predetermined from some bird's eye view of the field. But generally, the strategy was to avoid idiosyncrasy: the pointers in a blind entry should point somewhere useful. There is nothing more alienating than, for example, to seek the definition of hirsute in a dictionary and find it as a blind entry pointing to nonglabrous. Now what?
The second edition differs from the first in several substantial ways with respect to presentation of material. Latinate signals to cross-references (q.v.) were replaced by See (a direct instruction to the user to consult another term; See also sends the user to related or alternate technical material), and all cross-references were grouped and located at natural breaking points in entries. Cross-references among language lists are indicated by small capitals in the opening paragraphs. Descriptive headnotes to composite entries in the first edition (“This article is concerned with …”) were trimmed markedly to schematic tables of contents. Italics, small capitals, boldface, and other typographical distinctions were simplified to reliance on italics only for citation and emphasis. These changes were made to promote ease of access and recognizability.
Personnel and procedures
All Topic Editors from the first edition were invited to oversee revision of their areas. Almost all agreed, and the reconstituted group was renamed Consulting Editors. All Editorial Advisers of the first edition were also invited to participate again, with most agreeing, and this reconstituted group was renamed Senior International Advisers. William Bright served as Senior Consulting Editor.
Consulting Editors were charged with determining a revision plan for their areas, which was discussed in detail via e-mail, telephone, and personal visits with the Editor in Chief. This plan had to estimate extent of revisions and recommend additions or deletions. Consulting Editors also contacted all authors and oversaw the submission of articles from contract to final editing. Articles by deceased authors had to be revised by new authors, and authors who had left the field had to be contacted to determine their role in the second edition. These personnel matters were labor intensive and time consuming, and the Consulting Editors rose admirably to the challenge.
Senior International Advisers offered commentary on the overall structure and coverage of the Encyclopedia. The Senior Consulting Editor planned the revision with the Editor in Chief in a series of meetings over several years and served as periodic troubleshooter throughout the emergence of the work.
Revision plans, procedures for revising articles or submitting new ones, and various notices about policies were posted on the Encyclopedia website, which was located on a server at the University of Delaware, the Editor in Chief's previous institution (http://www.udel.edu/billf/iel.html). The Internet proved invaluable in this project since it allowed not only rapid transmittal of large amounts of information in various formats, but also remote, asynchronous access to a variety of materials for all contributors.
The Encyclopedia is a complicated work, with a wide range of material illustrating and exemplifying issues and arguments. All this material was extensively revised. New formal notation was added and maps were redrawn. The boundaries of polities have changed markedly since 1992, as have the locales of languages within and across those polities. It is December 2002 as this introduction is being written: imagine how Central Asia looked in 1992! Color images of brain functioning were added. Tree diagrams of both linguistic structure and language families were edited. Diagrams of theoretical models were emended, as were orthographies. Linguistics, it turns out, is a heavily visual discipline.
Articles on language families are followed by lists of languages. IEL has a range of articles on languages, from extensive treatments of particular languages (“Zulu”) to long expositions on major language families (“Germanic Languages”). There are also paragraph-length articles on intermediate-level language families (for example, “East Fijian Languages”), which were written by Bernard Comrie from Barbara Grimes's data and are included only if they appear as nodes in the Ethnologue database. These lists were compiled by Barbara Grimes— not by authors of the articles— using the Ethnologue and databases of SIL International (formerly the Summer Institute of Linguistics). These lists also include updated information on the demographics of the speakers. There remain great controversies in the field over which languages belong to which families, and, indeed, some of the groupings in the lists are at odds with the positions of the authors of the articles. The goal of including the lists was not to resolve controversies— or promote them!— but to ensure that the user has maximum information.
Many more short biographies of figures in the history of linguistics were added to the new edition for the sake of comprehensiveness. So biographical entries now appear for figures from Aristotle and Joseph Justus Scaliger to Lev Vygotsky and Ken Hale. The motivation behind expanding these biographies was to offer the user a kind of dramatis personae of the field and so make the Encyclopedia more like a handbook. Most of the biographies were adapted from The Concise Oxford Dictionary of Linguistics, edited by P. H. Matthews, and are reprinted here courtesy of Oxford University Press.
The glossary that appeared in the first edition was eliminated and replaced by a more detailed index, supplemented by careful attention to definition of key terms at points where these terms initially appear in articles. It was thought that an extensively revised glossary, coordinated with the text of the volumes, would lead to duplication. Backmatter also includes a new systematic outline and a directory of contributors.
I must thank Bill Bright and Oxford University Press for giving me the opportunity to oversee this work. I know I drove Oxford's editorial staff to exasperation with some of my worries and insistence, but I hope not to fatal lengths. Their patience and diligence, as well as Bill's, are models to us all. The Consulting Editors put up with my polite reminders and incessant hectoring. Stephanie Baker, my very capable research assistant, read almost the entire work to check for ease of exposition and to ensure definition of terms. Bill Idsardi often gave me excellent advice on technical matters.
As the Encyclopedia was coming to the final phase of production, I left the University of Delaware, where I had been for twenty-three years, to become Dean of Arts and Sciences at George Washington University. This change, of course, was equivalent to adding on six or seven more encyclopedias to be completed at the same time! I thank my colleagues and staff at UD for their help and my new staff at GWU for seeing me through these trying times.
As always my wife, Maria, my children, Christopher and Emma, and our array of cats (Chloe, Pierre, Jasper, Lionel, Zeke, and Maisie) provided unyielding comfort and stability. Eternal thanks to them.
William Frawley, Editor in Chief