Papers of the Fortieth Algonquian Conference: Actes du Congrès des Algonquinistes

Cree Syllabic Fonts: Development, Compatibility, and Usage in the Digital World

Cree Syllabic Fonts: Development, Compatibility, and Usage in the Digital World Bill Jancewicz and Marie-Odile Junker SIL International and Carleton University intrOductiOn Like other minority languages, but maybe even more so, Aboriginal languages are facing challenges in encountering information technology (henceforth IT). Our experience in helping develop resources for typing in Cree syllabics in the IT area (see also Jancewicz and Junker 2002) has led us to explore the following questions: How do changes in IT affect languages such as Cree? • What are the best IT tools for promoting and preserving the language? • Can these tools be understood and accessed by the people who need them? • To what extent are minority languages vulnerable with respect to IT? We will start by discussing the recent history of character encodings and the effect it has had on Cree users. We then examine current issues pertaining to display of fonts, keyboarding, conversions, and distribution. Finally, we discuss some current collaborative IT applications and practices involving the Cree language within the East Cree community of speakers.1 151 1. We wish to thank all the Cree writers, speakers, and linguists who have participated in our dialogue about Cree fonts over the year, as well as the audience at the 40th Algonquian Conference, Minneapolis, MN, October, 2008. Special thanks to Timothy di Leo Browne for editorial comments, and to Delasie Torkornoo for technical support. Research for this paper was partially funded by a SSHRC grant (# 856-2004-1028). 152 Bill Jancewicz and Marie-Odile Junker FrOM legacy (8-Bit) encOdings tO unicOde The history of the development of computer technology in the 1980s and 1990s provides the reasons for some of the hurdles that needed to be overcome in order to provide efficient usability of Cree syllabics on computers. When access to personal computers was first within reach of Cree speakers, the physical limitations of the computer’s machine language forced font designers to independently develop their own encoding systems. Within the 8-bit/1-byte framework there is a theoretical limit of 256 characters. But because certain characters were reserved by the operating systems and the application programs, users were left with a practical limitation of fewer than 200 characters. The standard “English” (so-called “Latin”) character set, with upper- and lower-case characters, numerals, and a few symbols, used up most of the character codes available in the first 128 characters that a standard keyboard could type. An additional set of “extended” (ANSI) characters was made available when Microsoft Windows and the Macintosh Classic operating systems were introduced, but unfortunately Microsoft and Apple did not agree upon which characters would be represented in their “extended” character sets. To make matters worse, Microsoft Windows and Macintosh also developed different approaches to the “code-page” (so-called “standard” re-coding of the extended character set to enable the use of some majority languages). By the mid-1990s the only workable solution for Cree language communities was to restrict the choice of operating systems used by the community, that is, to require everyone to use either Windows or Macintosh, but not both. Moreover, a computer technician or font designer had to be retained to develop a “hacked” font that contained the characters that the language community needed. Having made the operating system choice, all the users then had to be trained in the use of this special local font. A computer technician or programmer was also tasked with converting previously keyboarded documents into the accepted local standard. Sharing documented across operating systems, or even across application programs, meant that elaborate conversion procedures had to be followed in order to maintain the integrity of even the simplest documents. In addition, many of the legacy (8-bit) fonts experienced interference problems with the AutoCorrect features of word processors. For example, features that automatically capitalized letters after periods, corrected “two initial capitals,” or replaced straight quotes with “smart” quotes often 153 Cree SyllabiC FontS caused the wrong syllabic character to be placed in context. Users routinely reported these problems without being able to identify their source. We have heard of people who were using a strategy of leaving blank spaces for the misbehaving character in their electronic document, then filling them in later by hand after printing. If such texts are retrieved from electronic archives in the future, they will have missing characters for this odd reason. For legacy fonts, all AutoCorrect features have to be...