Languages in the UNC-Chapel Hill Libraries

Introduction
First steps
More focused searching
Using language(s) in word processing and on the internet
Internet language resources

Introduction

As a major institution of higher education, the University of North Carolina at Chapel Hill offers instruction in a number of lesser-taught languages, such as Arabic, Hindi/Urdu, Indonesian, Kiswahili, Vietnamese, and Yucatec Maya, in addition to the more standard array of Chinese, French, German, Greek, Latin, Spanish, and so on. The Libraries' mission is to support the teaching efforts of the University, and it has established strong collections in various languages, both in the sense of collecting materials about languages, as well as materials in different languages about other topics. UNC-Chapel Hill has further extended its language collections through close cooperation with the library system at Duke.

• framing the request:
  

  "I'm looking for material in [fill in language of choice]" in the context of the library collections is not as obvious as it might first appear. Frequently, this request can be more precisely framed in one of two ways:

  1. "I'm looking for works in [fill in language of choice] literature":
     

     The results of a search, say, for Spanish literature would obviously include works by the likes of Cervantes, Octavio Paz, and Isabella Allende. This will include the works of these (and other similar) authors, whatever the language in which a specific edition has been published. The tales of Cervantes' Don Quijote have, of course, been translated into many languages, which is to say, that though it remains a work of "Spanish literature," it may appear in a language other than Spanish.

  2. "I'm looking for a book on algebra, but I'd like it in [fill in language of choice]":
     

     Supporting the research needs of a diverse population, the UNC-Chapel Hill library system has collected materials in many languages. Materials focused on art, business, history, technology, and so on, can appear in any language.

• limiting a catalog search by language:
  

  There are two interfaces to the electronic catalog, and both allow for limiting an author, title, or subject search by language. The web version offers a "drop down" box on the search page, and this box provides the language options from which a choice can be made.

  A search in the telnet version can be limited in two ways. If the coding is known (and there is a complete listing of such codes at http://lcweb.loc.gov/marc/languages/ on the Library of Congress website), the search can be limited from the start, e.g., "t=king lear|la=fre" will find copies of Shakespeare's King Lear in French. If the coding is not known, a regular search can be entered. Once the results appear, the next step is to enter "lim" and the catalog will then prompt the user through a series of limiting choices.

• Library of Congress classifications:
  

  The UNC-Chapel Hill library system employs the Library of Congress classification to arrange its materials by subjects. This system places books about languages in certain call number ranges. Note that the following is only a brief outline. More details can be found in the LC Classification Outline, available in print at the Davis Library Reference Desk, or on the Library of Congress website at http://lcweb.loc.gov/catdir/cpso/lcco/lcco.html (this link requires Adobe Acrobat).

  • P: works on language in general
  • PA: Classical languages and literatures [this includes Latin and Greek]
  • PB: Celtic languages and literatures [this includes Irish and Welsh]
  • PC: Romance languages [this includes French, Italian, and Spanish]
  • PD: Germanic languages [this includes the Nordic languages]
  • PE: English language
  • PF: West Germanic languages [this includes German and Dutch]
  • PG: Slavic, Baltic, and Albanian languages [this includes Polish and Russian]
  • PH: Finno-Ugric and Basque languages [this includes Finnish and Hungarian]
  • PJ: Oriental languages and literatures [this includes Arabic and Hebrew]
  • PK: Indo-Iranian, Armenian, and Caucasian languages [this includes Sanskrit, Hindi, and Urdu]
  • PL: languages and literatures of Eastern Asia, Africa, and Oceania [this includes Chinese and Japanese]
  • PM: American Indian languages

  These classifications are for materials about languages (and, in a few -- although not all -- cases, about the literature originally written in those languages). Again, materials about other topics, whatever the language in which they are printed or published, will simply appear in classification areas pertaining to those topics.

• initial articles in titles:

  There are currently two catalogs for UNC-Chapel Hill materials, one electronic (containing some 80% of holdings) and then the remnant of the old card catalog (containing almost 20% of holdings). As already noted, there are two interfaces for accessing the electronic catalog -- web and telnet. One rule holds true for all versions of the catalog: when searching by title, any initial article -- whatever the language -- is dropped, although any internal articles are retained. Thus, titles are alphabetized without reference to opening words like

  
  
  • the English the, a, an
  • the Danish den, det, de, en, et
  • the Dutch/Flemish de, het, 't, een, eene
  • the French l', le, la, les, un, une
  • the German der, die, das, ein, eine
  • the Hungarian a, az, egy
  • the Italian il, lo, i, gl', gli, la, le, l', un, uno, una
  • the Norwegian den, det, de, dei, en, ein, et, ei, e, eit
  • the Portuguese o, a, os, as, um, uma
  • the Romanian l, le, un, o
  • the Spanish el, los, la, las, un, unos, una, unas
  • the Swedish den, det, de, en, ett
  
  

  This practice includes articles that may be attached directly to the nouns they modify, as in the Hebrew "ha-" or the Arabic "al-". Again, this is the case only when these words appear as the first word in the title. Should they appear anywhere else, e.g., Journal of the American Medical Association, they must be included in the search.

• "modified characters" in the catalog:
  

  Many languages employ some version of the Roman alphabet which is used in English, although with occasional modified forms in order to accommodate sounds that do not appear in English. These modified forms, e.g., letters with accents, umlauts, or other diacritical marks, may or may not require special treatment in the library setting.

  In the case of the card catalog, modified characters are for the most part treated as if they were unmodified, e.g., "è" and "é" are filed as if they were simply "e" alone. The exception to this practice is the treatment of modified letters in the Germanic languages, which are filed as if they were "spelled out": "å"="aa," "ä"="ae," "ö"="oe," "ü"="ue," "ø"="oe," and the Dutch "ÿ"="ij." Cards may be printed in either fashion, e.g., with either an "ä" or an "ae," but they will be interfiled as "ae."

  The electronic versions of the catalog ignore diacritical marks in the search (and, in fact, will likely misread any attempt to type with an accent or an umlaut), but make the attempt to reproduce them when displaying the records retrieved. Thus, when searching for an item with accents, umlauts, circumflexes, etc., an "è" is rendered as "e," an "ö" is typed in as "o," and so on. The web version of the electronic catalog tends to display such modified letters better than does the telnet version, which may need to resort to completely different lettering because it fails to "understand" the underlying code properly. This may mean that a retrieved record will "look funny," although it still refers to the correct item. Note that if an item, most probably in one of the Germanic languages, had been electronically cataloged with the modified letter(s) "spelled out," it can only be retrieved this same way. This is likely an occasional problem and only with older materials.

• transliteration(s) in the catalog:
  

  Other languages employ completely different writing systems. These systems must then be reworked or "transliterated" into the characters of the Roman alphabet (with or without modifications) in order to be used in the library catalog.

  Such transliterations may prove tricky, as there may be more than one acceptable way to make such changes. The "official" transliteration options chosen by the Library of Congress (and thus by the UNC-Chapel Hill library system) can be found in ALA-LC Romanization Tables, the 1997 edition of which can be found at the Davis Library Reference Desk (call number P226.A4 1997).

  Because scholarship changes over time, such "official" standards may also change. Currently, the Library of Congress is engaged in a changeover from one system of transliterating Chinese characters to another. This will make a major difference in searching for materials in Chinese. To help explain these changes, there is a report available on the Library of Congress website at http://www.loc.gov/catdir/pinyin/pinyin.html.

• dictionaries:
  

  The catalogs can be searched for basic dictionaries in specific languages by employing the standard subject search formula *xyz language -- dictionaries* [note: in this and all following suggested catalog subject searches, the term(s) between the * * should be typed into the catalog, without the * *], in which the "xyz" stands in for the language of choice, e.g., *French language -- dictionaries*. This search will include single language dictionaries, i.e., a French language dictionary that provides descriptions of French words in French.

  At times what is desired is a bilingual dictionary, e.g., French-English. The standard formula can be modified to provide appropriate records by adding the additional information, e.g., *French language -- dictionaries -- English*. In the case of a multi-language dictionary, the subject search would be altered to *French language -- dictionaries -- polyglot*. This last would provide records for works like a French-Basque-Spanish dictionary.

  Language dictionaries with a specific linguistic focus or on a special language topic might be cataloged somewhat differently. For example, dictionaries that focus on the etymology of Italian words can be isolated by searching for *Italian language -- etymology -- dictionaries*.

  It is also possible to start from the topic or subject area and then append "dictionaries" to find works that help define that topic, e.g., *technology -- dictionaries*. This can be further modified to include a specific language, e.g., *technology -- dictionaries -- Czech*. Such topics include, but are not limited to, the following: business, computers, finance, philosophy, and printing.

• grammars, verb tables, etc.:
  

  In addition to dictionaries, there are other types of books about languages. For example, grammars of Latin can be found by searching for *Latin language -- grammars*, while works concerned with the sounds of Greek are under *Greek language -- phonology*, those concerned with the history of Sanskrit are under *Sanskrit language -- history*, and those concerned with Chinese dialects are under *Chinese language -- dialects*.

  The popular language aids that fully conjugate either 201 or 501 verbs in a given language, e.g., 201 Hebrew Verbs Fully Conjugated..., can be found by searching with the usual pattern -- appropriately modified: *Hebrew language -- verb -- tables*.

• learning/teaching materials:
  

  Materials for learning a language can be searched in several ways, e.g., *Spanish language -- self instruction*, *Russian language -- textbooks for foreign speakers*, or *Arabic language -- readers*, while materials for teaching a language might be searched as, for example, *German language -- study and teaching*. These can all be further specified if there is a target language in mind, e.g., *Swahili language -- textbooks for foreign speakers -- Norwegian*.

  The printed word is not the only medium for learning languages and the UNC-Chapel Hill library system also offers relevant materials in other formats. The Media Resources Center houses such materials, which can be searched by limiting a subject by specific formats, e.g., the subject search *Spanish language*, if limited to "recordings (spoken word)", will return such titles as Living Spanish, which is a two-CD set. The "audio-visual" format may also return relevant titles. These limits can be found in a drop-box form on the author, title, subject search page in the web version, or can be activated in the telnet version by following up an author, title, or subject search by entering "lim" (this will then prompt a user with a series of limiting choices).

  There is a language lab on the lowest floor of Dey Hall, but its use is restricted to students enrolled in language classes.

• computers in general:

  Everything in computers can be reduced ultimately to more or less established codes. The characters that are used to represent the alphabet are no exception: every letter has its own code, which software is programmed to recognize.

  The basic Roman alphabet has a long-established, and more importantly, a long-accepted code. Modified characters can be handled in the exact same fashion, although because the degrees of establishment and acceptance may not be as strong, different operating systems and different programs may begin to diverge. It is possible that occasional codes will not be understood in the same way by all programs; the result is that the desired character may not appear. This problem is greater the further the departure from an unmodified Roman alphabet.

  Available modified characters -- and sometimes even characters from non-Roman alphabets -- can be found in a "character map." In Microsoft Windows products, this can be found by opening up the "start menu," choosing "programs," and then "accessories." Opening up the "character map" reveals a chart -- sometimes, more than one -- of standard and non-standard characters and the coding used by these characters. These may be added, e.g., to a word processing document, by using the "ALT" key and the coding provided, e.g., the "ä" can be used by typing "ALT" and the number 0228. It may be possible to develop "shortcuts" if such modified characters are used frequently.

• computers in Davis Reference:
  

  When computers are set up, certain defaults are established. These defaults include standard language of operation. The computers in Davis Library almost always default to an American English standard, which clearly privileges the Roman alphabet along with some basic modified characters, although it is possible to establish other standards, e.g., Cyrillic. A few computers in the public area of the Davis Reference Department have been set up with the possibility of an alternative language default (although this default may need to be "activated" on a session-by-session basis). These machines are as follows:

  • INET #5: Chinese, Japanese, Korean, Cyrillic-based languages
  • INET #6: Chinese, Japanese, Korean, Cyrillic-based languages
  
  
• using language(s) on the internet:
  

  When working with the world wide web, say with a web browser like Netscape or Internet Explorer, the words that appear on screen may be either graphic or textual in nature. There should be no difficulty in the browser's ability to show graphics correctly (even if the language uses a non-Roman alphabet), because the computer is merely showing a picture that happens to take the form of letters or characters. This is why, for example, when viewing a non-English language page, the headline or the masthead of the webpage may appear legible, although the actual body of text seems garbled.

  Text is different. First, the webpage must be coded correctly, i.e., according to the widest, currently-accepted rules. Such appropriate HTML coding can be found among the fairly-authoritative webpages of the World Wide Web Consortium (the homepage of which can be found at http://www.w3.org). This consortium provides information on the coding appropriate for various letters, alphabets, and writing systems (online at http://www.w3.org/International/).

  Second, individual browsers may need to be "instructed" to understand such HTML coding, if something other than the usual default is being used. As noted above, INET machines number 5 and 6 within the Davis Library Reference Department offer the flexibility needed for enabling the reading of materials in other languages. Outside of the library system, it may be possible to download the necessary fonts, either from commercial or noncommercial websites.

The internet has expanded the concept of a library's holdings, at least in so far as it enables patrons to access materials beyond a library's physical location. Here, then, are some relevant language resources to be found on the world wide web.

• Looking for an online dictionary? Chances are good that it can be found (along with grammars and other language-related things) at http://www.m-w.com/
   
  
• Looking for online links related to specific languages? Try the iLoveLanguages website, which was until recently the Human-Languages Page (online at http://www.ilovelanguages.com/).
   
  
• The "Languages" section of Yahoo! can be found at http://dir.yahoo.com/Social_Science/Linguistics_and_Human_Languages/Languages/
   
  
• There is, among other links, a section for "Language Resources" on the Linguist List website at http://www.linguistlist.org/
   
  
• As an example of East Asian language resources, http://www.zhongwen.com/ offers a website on Chinese characters and culture, with links to texts and news.
   
  
• electronic texts
  • The Electronic Text Service of the Columbia University library system provides a listing of online text collections organized by language at http://www.columbia.edu/cu/lweb/indiv/ets/offsite.language.html
  • 
    
  • The Western European Studies Section of the Association of College and Research Libraries offers WessWeb, a set of links to electronic text collections of Western European literature. It is found at http://www.lib.virginia.edu/wess/etexts.html

For more Arts and Humanities Reference assistance,
Please contact Tom Nixon, Humanities Librarian
tommy_nixon@unc.edu
(919) 962-1151

Can't Find What You Are Looking For? How Can We Help You?