１ Complexities of Ethiopian Sign Language Contact Phenomena & Implications for AAU Michael W Morgan It is possible to fail in many ways...while to succeed is possible only in one way. (Aristotle, Nicomachean Ethics)1 Abstract: This paper provides a “tour” through the gamut of Ethiopian Sign Language language contact phenomena and patterns manifested in the Ethiopian Sign Language and Deaf Studies program administered under the Department of Linguistics at Addis Ababa University. While a very small and limited microcosm, there is every reason to believe that most of the phenomena and patterns observed in this program might be fruitfully applied, mutatis mutandis, to the macrocosm of Deaf Ethiopia at large. And, as these observations speak to important larger issues, these observations are accompanied by commentary on why these manifestations may NOT be what they seem, and how they might conflict with the vision statement of the Ethiopian Sign Language & Deaf Studies program at AAU. 1. Introduction Ethiopia is a country with some 85 living languages (Lewis 2009), and so language contact is to be expected. Language contact can have many results: borrowing (whether simple lexical loan words or more complex structural borrowing), bi- and multi-lingualism (at both individual and community levels), code-switching and code-mixing, and so on … sometimes leading finally to a shift in language identity, loyalty and use, and-- in the worst case -- ultimately to language death. Ethiopian Sign Language (hereafter: EthSL, although abbreviations such as ESL and YEMQ are used elsewhere) -- at least as seen from the Ethiopian Sign Language and Deaf Studies program (hereafter: “the ESLDS program”) at Addis Ababa University (AAU) -- evidences all of these results (and others), except, fortunately, not language death … not yet, at least. However, there appears to be some cause for concern. 2. Background of EthSL in Ethiopia. Is EthSL Endangered? UNESCO recognizes a number of criteria which qualify a language as endangered. The most obvious of these might be (a) absolute numbers of speakers, but other factors such as (b) proportion of speakers within the total population, (c) intergenerational language transmission, (d) community members’ attitudes towards their own language, (e) shifts in domains of language use, (f) governmental and institutional language attitudes and policies, (g) availability of materials for language education and literacy, (h) response to new domains and media, and (i) type and quality of documentation are also vital to language vitality (UNESCO Ad Hoc Expert Group on Endangered Languages 2003). There is no doubt that a fair number of the 85 living languages in Ethiopia are at risk under criteria (a). It is unclear exactly what the absolute numbers of speakers of EthSL are. According to the most recent Ethnologue, it counts 1,000,000 “speakers”2 (Lewis 2009, citing Ethiopian National Association of the Deaf, 2005). This would mean that EthSL has more “speakers” than all but eleven (i.e. 13%) of the 85 living languages recorded for Ethiopia.3 On the other hand, as cited in the overview of the ESLDS program, “[a]ccording to the 1994 Housing and Population Census of Ethiopia, there were 190,220 deaf and hard of hearing people in the country” (Department of Linguistics n.d.). Even at this much lower figure, EthSL would have more speakers than 76% of Ethiopian languages. Unfortunately, neither a simple calculation of (a) absolute numbers nor (b) proportion of speakers within the total population gives an accurate picture of the vitality of EthSL, for there is one demographic characteristic of EthSL which distinguishes it from all other languages in Ethiopia. The difference can be seen if we compare EthSL with another Ethiopian language having similar numbers of speakers (choosing for this purpose the lower, Housing and Population Census, numbers for EthSL as a “worst case scenario”). We can take Konso with its 195,000 speakers as  ２ our comparison (all data here and below are from Lewis 2009). While both EthSL and Konso have similar numbers of speakers, and thus represent a similar proportion of the total Ethiopian population, there are reasons to believe that the numbers for EthSL represent a much lower degree of vitality. While Konso (and thus also EthSL) speakers represent only 0.25% of the total Ethiopian population, within its own home domain (“south of Lake Chamo, Sagan River bend”), the proportion of Konso speakers relative to the total local population is much larger, which is also reflected in the high number of mono-linguals for Konso (138,696, or 71%). EthSL, on the other hand, has no home domain, and the deaf population is spread (and thus diluted) throughout the country.4 Even at the level of an individual deaf person’s life, few domains of deaf life in Ethiopia are dominated by EthSL. Most deaf children are born into all hearing (and thus non-signing) families, and most adult deaf have hearing children and many perhaps also hearing or hard of hearing (as opposed to deaf) spouses, so not even the family is a safe signing domain. This also means that there is relatively little natural intergenerational transmission within the family , and so EthSL is vulnerable also under UNESCO’s criteria (c). Although there are no figures on numbers of monolingual EthSL speakers, it is unlikely that the percentage is anywhere near half, not because deaf are bilingual (in the real world an enviable thing, even if it increases the risk of language shift), but rather because so many of them are multiply semi-linguals: raised orally, or with limited and imperfect sign input at best, given little and poor education / literacy training, and so not fully competent either in sign language nor in any spoken or written language. Although under past regimes a single language, Amharic, was privileged and used exclusively in most official public domains, under the current constitution of the Federal Republic of Ethiopia Article 5.1 “[a]ll Ethiopian languages shall enjoy equal state recognition” (Constitution of the Federal Democratic Republic of Ethiopia, 21 August 1995), and, deaf express concern that EthSL is not specifically referred to, neither is any other language and so this need not necessarily be a concern, as it would fall under the purvey of “all languages”. Thus, EthSL need not be endangered under UNESCO’s criteria (f), at least in terms of the national governmental law/policy. However, policy and practice often do not see eye to eye, and it is not only a question of “equal state recognition” but rather of what policies and practices are implemented to see that they get equal state treatment. Recently in Ethiopia, there have been some encouraging policies and practices. For example, there have been some significant expansion in the range of domains of sign language use (criteria e); sign language has found its way into television news, and it has found its way, in the form of “inclusive education” into “normal” public schools (in addition to deaf schools, which in Ethiopia continue to be private, church run). Finally, the ESLDS BA program here under review is also such a new domain: higher education.5 Against this background, it would seem that the opening of a BA-level ESLDS program administered by the Linguistics Department at Addis Ababa University would be an unquestionably welcome measure to counter the endangerment: not only expanding (and elevating) the domain of sign language use, but also a place for production of quality documentation of the language (criteria i) and educational materials (criteria g), and a positive influence on deaf community attitudes towards sign language (criteria d). Unfortunately, although the ESLDS program could in fact be doing all these things6, as we shall see below, it is also doing something else: it is diluting the status of EthSL by failing to distinguish EthSL from its “competition”, and thus “aiding and abetting the enemy”. 3. The ESLDS Program: Linguistic Demographics As this paper describes the language situation in the ESLDS program during the 2009-2010 academic year, a brief summary of the linguistic demographics of the program is in order. The student body of the program during this period included two cohorts: 25 third-year students and approximately 40 second-year students.7 Of the third-year students approximately three quarters are deaf8. The percentage drops to half or less among the second-year students. Among both deaf and hearing students, there are members who knew no sign language when they entered the  ３ program – although of course those numbers were much higher among hearing students. Overall, the number of students who were proficient EthSL speakers when they started the program is significantly less than half. Other core members of the ESLDS community include the teaching staff and interpreters. All full-time teaching staff, all interpreters, and all part-time and external subject teaching staff except one are hearing. At the beginning of the year all three full-time teaching staff were signers: two of them were recently arrived ex-pats, and thus fluent in sign languages9 other than EthSL (though competence in EthSL gradually improved throughout the year). The other full-time teaching staff (who is also the program coordinator) is an L2 signer of EthSL and/or Signed Amharic/English. A fourth hearing non-signing full-time staff member was added halfway through the year. The two ex-pat lecturers preferred to lecture in sign language without any accompanying oral speech, while the third full-time lecturer demonstrated a preference for a mixture of means of communication10. All courses taught by non-signers were conducted with the aid of the interpreters, who varied considerably in their competence in sign language – from extremely poor to moderately competent (as judged by the reported ability of deaf signing students to understand the content of their interpreting of lectures), and also in whether they typically signed EthSL or Signed Amharic/English or some mix. Students also had contact with other members of the AAU academic and non-academic community, and with the exception of the handful of Deaf students, and a few hearing students outside the ESLDS program who had acquired a certain amount of EthSL, all members of this extra-curricular community are hearing and non-signing. 4. Idealized View of Languages in Contact (and in Use) in the ESLDS Program It is dangerous to hypothesize exactly what image the founders and administrators of the program had in mind when they inaugurated it, but we can perhaps – based among other things on the description of the program and the curriculum (Department of Linguistics n.d.) – imagine that it looked something like the following: Most importantly, the diagram illustrates three things: (a) the relatively limited number of competing languages (four); (b) the fact that clearly two types of language are involved: sign languages to the left of the vertical line, and “spoken” languages, (albeit, presumably, in written form) to the right); and (c) the fact that the languages involved are all clearly distinct, discrete, well defined (all except perhaps EthSL, that is), and easily distinguishable; no one would, for example, be misled into thinking that a person speaking English was in fact speaking Amharic. In the center of this diagram are two languages: EthSL and English. As the name of the program suggests, EthSL is envisaged as being central to the program, with all classroom lectures and discussion made, in theory, available to Deaf students through sign language (either by a signing lecturer or via sign language interpretation). English, on the other hand, is the official language of AAU, and also the language of all written material – be that study materials, supplemental readings, homework, papers, or examinations. Slightly outside this center, and above, is American Sign Language (abbreviated ASL below). For a variety of reasons (reasons which we shall see below are both mistaken and misleading), ASL is seen to have some important relation to the course of study, and so was both included in the  ４ curriculum (as a single course, not to be found on the online curriculum, and so possibly subsequently deleted), and also fluency specifically in ASL was listed as a preferred qualification when ex-pat teachers were recruited (Department of Linguistics 2009). Finally, although having no official status within the faculties of the university, Amharic is the de facto operating language, not only of AAU, but of Addis Ababa and Ethiopia as a whole, and also the schools which most the students enrolled in the program would have gone through prior to entering the program. 5. More Realistic View of Languages in Contact (and in Use) in the ESLDS Program The diagram presented above – and thus the presumed vision of the language situation within the ESLDS program when it was established – is clearly oversimplified and misleading. Instead, we can substitute the following (which, granted, is still somewhat oversimplified): In this diagram, EthSL and English still occupy the center, supported by Amharic and ASL, as in the first diagram. However, the number – a confusing array one might even say – of languages multiplying in the periphery easily distract attention away from this center. Additionally, although the modified diagram continues to project the image of discrete language entities divided by the vertical line into signed and spoken, this is clearly not a true reflection of reality. Let us examine a few of the more important complications in turn: 5.1. EthSL, and What Has ASL Got to Do With It? Although not “politically correct”, in fact it has not yet been established that there is a single entity we can call “Ethiopian Sign Language”. As the Ethnologue notes “[s]everal sign languages are used in different schools for the deaf” (Lewis 2009), and so it is probable that several sign languages also exist in the Deaf community. Certainly there are a range of varieties used by students even just within the ESLDS program.11 It is thus unclear whether the “Ethiopian Sign language” being promoted within the program is in fact a single real entity, a connected continuum of sign language varieties, or an artificial construct (which then can serve, once codified and forced on the deaf community, as a national sign language). Often it appears that there is no clear idea within the administration of the program of what “Ethiopian Sign Language” is, and it is assumed that it is one “language”, without that language ever having been defined or its existence even established, and with in fact a number of other “languages” being practiced and promoted equally, and all called by the same name.12  ５ Interestingly, several features which Woodward (1994[1973]) notes as being indicative of non- native ASL signing are quite common in signed classroom discourse in the ESLDS program. While this may derive, as it does in the American context, from a shift along the sign language-signed language continuum (and thus such signs as VERY and AND 13 are a part of the signed language in Ethiopia as in America), it is also possible that claiming that EthSL is based – in any way and to any degree – directly on ASL is mistaken. It is not ASL that came to Ethiopia with the establishment of Deaf schools by American (hearing) missionaries, but rather Signed English. Signs from Signed English have been absorbed into EthSL14, but also competence in Sign English has been – to some extent at least – acquired by many if not most who attended deaf school. And it is natural that such Signed English would be one resource used by deaf signers when they encounter and attempt to communicate with hearing people, even when those hearing people are themselves fluent signers. 5.2. EthSL and Signed Amharic / Signed English, TC (Total Communication) , Sim-Com (Simultaneous Communication), Pidgin Signing, and All the Rest There is a tendency, frequent among people who are not sign linguists and not infrequent among even less nuanced sign linguists, to refer to all communication using signs as “sign language”. In fact, however, there is a wide – and potentially confusing – variety of non-language varieties involved, and all of these are to be seen in the day-to-day life of the ESLDS program. As perhaps half of the students are not fluent signers when they enter the program, L2 signing is one such variety. As there are no course books in EthSL grammar, and thus students (in the first cohort at least) may have had no idea of a “target language”, some of this L2 signing might properly be called “pidgin signing” – or “signing without any noticeable grammar”. Third-year students appeared unable to distinguish even between the best L1 and the worst L2 signing.15 Two other varieties entering the mix are Signed Amharic and Signed English, which have already been referred to above. In Signed Amharic (mutatis mutandis, Signed English) signs from the natural deaf sign language (i.e. EthSL) are used, supplemented by artificial signs (often the superimposing of a finger spelling hand-shape of the initial letter of the written word onto a natural sign), signed in the order of the spoken language. Although potentially supplemented by made-up signs for grammatical units (and the total then called “Signed Exact Amharic”), typically grammatical elements – both of the spoken language and of the sign language – are omitted and/or inconsistently used.16 Such signing is typical when the signer is also voicing the spoken language, or when a less proficient hearing signer signs while thinking in the spoken language. Signing while speaking at the same time is called simultaneous communication, or sim-com (and often mistakenly called Total Communication, which it is not). Needless to say, unless the spoken language is an isolating language with no grammatical morphemes, much is lost in such signing, to the extent that it is often incomprehensible to deaf signers, as incomprehensible as spoken language – and which therefore qualifies as “nonsense”, or a mere waving about of one’s hands.17 Although none of these various forms of (attempted) communication is EthSL, confusion abounds. For example, the only published local linguistics research on EthSL (Eyasu 2005), while purporting in its title to be a treatment of “Copula Constructions in Ethiopian Sign language”, in fact, even a cursory look at the examples – aside from knowledge that EthSL has no copula! – would indicate that the topic of discussion is Signed Amharic and not EthSL. Similarly, in a PR piece about the program (Eyasu 2009) reference is made to Netzley (2005) as being a comparison between ASL and EthSL signs, whereas in fact that work clearly states that the source of the American signs used in the comparison is a “Signed English book”. 5.3. Contact Signing, International Signs and Mixed Signing A final set of varieties enter the mix as the ESLDS program had during this year not only two signing ex-pat lecturers but also, for one semester, a foreign deaf student. When deaf (or deaf-like) signers of distinct sign languages attempt to communicate, even if they have no previous knowledge of the sign language of the other person, something happens which is quite different from what happens when two hearing people with different languages meet. Although sign languages can be quite distinct in their lexicons and in the details of their morphology and syntax (as demonstrated in e.g. Zeshan 2006 and Zeshan & Perniss 2008), they share much in their “meta-structure”, and so  ６ negotiation of meaning is much easier. Unlike spoken pidgins, such contact sign languages typically have full grammars. Even International Signs, initially simply a word list without any guidance as to grammar, is typically produced with a complete grammar; the signer simply overlays International Signs onto his own native grammar, or to a compromise between that and the grammar of his interlocutor. Unlike Signed Amharic, such signing would perhaps qualify as a “mixed language” (i.e. mixed signing). 5.4. Code Sliding While it may already seem like a too confusing multitude of languages and varieties to be navigable, things are complicated by the fact that, within utterances, and within discourse, a single speaker may sign using more than one variety. While code switching and code mixing is indeed involved, often there is something I shall call “code sliding” which occurs. Take the case of a speaker who “mixes” EthSL and Signed Amharic. As has often been noted (e.g. Stokoe 1992[1960], Woodward 1994[1973], Lucas & Valli 1990), sign languages exist on a continuum with the signed form of the spoken language. So, for example, we can posit a continuum existing between EthSL and Signed Amharic. An individual utterance or segment of discourse need not be located clearly at one end of the continuum or the other; it need not be identifiable as clearly EthSL on the one hand or Signed Amharic on the other. Rather it may contain elements of each – in differing ratios – and thus be considered a mixing of languages. The complication comes from the fact that, while in spoken discourse code switching (etc) the switch is between forms which clearly belong to different codes, in sign discourse the switch may be between forms which belong to neither code, but rather to a hazy middle ground on the continuum. In code switching between Amharic and English, the switching or mixing is between forms which are clearly definable as Amharic and forms that are clearly definable as being English, while with sign language, the forms may be located anywhere along the continuum, and the “location” may be fluid back and forth, up and down the continuum within the discourse – what I call code sliding. Finally, the sliding may possibly occur along several continua at the same time, within the same course. Thus a speaker may “slide from more or less pure EthSL towards Signed Amharic, only to then slide in the direction of ASL – or gesture. And their interlocutor may do the same, but along a slightly differing mix of continua. 6. Conclusions: Implications for New Directions To paraphrase Aristotle, “all decisions are ethical decisions and all action is political action”. Within the context of a long history of audism and hearing hegemony in Deaf education, and of the current dangers of language endangerment outlined above, the establishing of this program must be seen as a (positive) ethical and political statement. However, having made an initial positive (albeit premature) step in the right direction does not assure that all subsequent implementations will automatically be positive. As Goethe said concerning marriage: “A confusion of the real with the ideal never goes unpunished". It would appear that the real language situation in the ESLDS program at AAU (and projecting from micro- to macrocosm, in Ethiopia as a whole) is orders of magnitude more complex (not only more confusing but also more confusable) than the ideal. Perhaps a change in mindset, a replacement of a simple ideal by a more messy and complicated reality will allow for corrective adjustments in the current implementation and administering of the program. Of course, more than a change in mindset will be necessary; collection and analysis of a wide range of data leading to convincingly believable and complete documentation of EthSL will be needed so that a defensive wall can be erected to keep at bay the horde of competing (and less valid) forms of simply waving ones hands around. Notes 1 This paper is intended not only as an attempt at “descriptive sociolinguistics”, but also a cautionary tale. Fledgling programs, such as the ESLDS program, are fragile. Not only can any number of things go wrong, but also, any one of those things – or any one person given responsibilities and power incommensurate with their level of knowledge and understanding –  ７ can do tremendous, sometimes irreparable, damage to the program, and to its credibility. 2 Use of terms such as “speaker”, “speak”, etc in this paper bear no relation to “oral speech”, but rather to any use of language, whether oral or manual. 3 It would also mean that EthSL has exactly the same number of speakers as the total Deaf population of the country (as given in the same source) – which is almost certainly not the case, not even if we define EthSL in the broadest of terms as “any and all forms of sign language”. 4 The ESLDS program prospectus (Department of Linguistics, n.d.) says in one sentence that most deaf and hard of hearing in Ethiopia are concentrated in urban areas, and in the very next sentence that the majority live in rural areas. Clearly, both cannot possibly be the case. Ultimately it does not matter for the argument here, even having them concentrated in urban areas would still probably mean that they are diluted across the various urban areas of Ethiopia. 5 All these positive signs are of course assuming that what is being used in these various domains is EthSL. Unfortunately, in almost every case this is not the case, and so the positive signs are just an illusion, or, if you are inclined to be cynical, little more than window dressing. We use the program at AAU as an example; in most cases, I fear, the status and situation of EthSL would fare even worse in the other domains listed. 6 I say “could be” because little actual documentation or materials development have been done, aside from linguistically unsophisticated lexicography projects – duplicating work being done elsewhere in the the country, all producing little more than uncontextualized word lists. 7 Although the ESLDS program is a three-year program, student status is officially referenced as if it were a four-year course. Thus, entering students are referred to as “second-year students” and students who are in their third and final year of the program are called “fourth year students”. Despite the confusion inherent in such nomenclature, I will abide by it here nonetheless. 8 Throughout this paper I will, for simplicity’s sake, often ignore the important distinction between deaf and hard of hearing. 9 Both in fact are near-native signers, and proficient in multiple sign languages. Experience in various deaf communities in multiple countries meant that, even though on arrival they knew little or no EthSL, this was not a major barrier to communication with the signing deaf students. Although mutual comprehensibility between unrelated sign languages is certainly higher than between unrelated spoken languages, other factors – including flexible co-operative communicativeness – are characteristic of deaf (and deaf-like) signers, so that “where there is a will, there is a way”. 10 For example, in the previous year, he had taught all classes orally, relying on (often less proficient) interpreters to sign for the deaf students. During the year here under review, although he did frequently sign without voice, he also often signed and spoke at the same time – a method called sim-com, described later in the paper. 11 The perception that varieties are more or less “ASL-like” may be as much a projection of competence in ASL, and the lack of analysis as to what other bases there may be for the differences observed. An extensive corpus has been recorded by the author and the other ex-pat lecturer, so that eventually this lack will be remedied. 12 This critique of the lack of a unified concept of what EthSL is should not be seen as an endorsement of attempts to eliminate (under the name of standardization) the variations inherent in EthSL, a variation natural to all languages, nor as a lack of appreciation for the incredible communicative flexibility of deaf signers the world over. Rather it is intended as a critique of the lack of rigor and direction in the program. 13 I follow common practice in sign linguistics literature of giving signs in small capitals. 14 Interestingly, students invariably identify signs as being either EthSL or ASL. Thus, for example, although a wide range of sociolinguistic variables were introduced and discussed by the teachers of “Sign Language and Society” course, when asked to present and explain examples of signs with variants, students typically described one (or rarely more) as being “Ethiopian” and the other(s) as being ASL. None were described as being from Signed English. Equally interesting is the fact that many of the signs identified as “Ethiopian” are in fact ASL, and some  ８ identified as ASL are in fact European (some clearly of Scandinavian in origin, although also many “International Signs”). 15 This was demonstrated when students were asked to compare videos of their retelling of a cartoon with those of five other students. A full half of the students, deaf and hearing alike, rated their own signing with the highest scores – including all but one of those students whom everyone else rated as worst. Clearly something more than just egotism is involved. 16 Numerous studies on Signed English in the ASL context indicate that not only do such methods not work in practice, they may well be inherently incapable of doing so even in theory (the dozens of articles reprinted in Gustason 1988; Johnson, Liddell & Erting 1989; etc). 17 Signed Amharic might – mistakenly – be thought of as a “mixed language”, with the vocabulary of EthSL and the grammar of Amharic. However, as indicated, the grammar – aside from mere word order – is typically missing or highly defective. Bibliography Department of Linguistics, Addis Ababa University (n.d.) BA in Ethiopian Sign Language. (accessed 1 May 2010: http://www.aau.edu.et/index.php/linguistics/ linguisticsprograms-/ba-in- ethiopian-sign-languages) Department of Linguistics, Addis Ababa University (5 February 2009) JOB POST: Lecturer in Sign Language Linguistics or Deaf Studies, Addis Ababa University, Ethiopia. Immediate [sic!] Vacancy Announcement. (accessed 1 May 2010 at: http://wecando.wordpress.com/2009/02/05/ job-post-lecturer-in-sign-language-linguistics-or-deaf-studies-addis-ababa-university-ethiopia/) Constitution of the Federal Democratic Republic of Ethiopia, 21 August 1995. (accessed 30 August 2010: http://www.unhcr.org/refworld/docid/3ae6b5a84.html) Eyasu Hailu (2005) Copula Construction in Ethiopian Sign Language. Unpublished MA thesis. Addis Ababa University. Eyasu Hailu (2009) Sign Language News at Addis Ababa University. (accessed 1 May 2010: http://www.signwriting.org/archive/docs6/sw0557-SignLanguage-AddisAbabaUniv.pdf) Lewis, M.P. (ed.) (2009) Ethnologue: Languages of the World, Sixteenth edition. Dallas, Tex.: SIL International (online version accessed 20 August 2010: http://www.ethnologue.com/) Hockett, C.F. (1960) The origin of speech. Scientific American, 203: 88-96. Johnson, R.E., Liddell, S.K. & Erting, C.J. (1989) Unlocking the Curriculum: Principles for Achieving Access in Deaf Education. (Gallaudet Research Institute Working Paper 89-3). Washington, DC: Gallaudet University Press. Lucas, C & Valli, C. (1990) ASL, English, and Contact Signing. In C. Lucas (ed.) Sign Language Research: Theoretical Issues. Washington DC: Gallaudet University Press. 288-307. Gustason, G., ed. (1988) Signing English, Exact Or Not: A Collection of Articles. Los Alamitos: Modern Signs Press. Netzley, J. (2005) Findings on the Deaf of Ethiopia and Ethiopian Sign Language. Unpublished ms. Stokoe, W. (1992) Sign Language Structure. Burtonsville, MD: Linstok Press. (reprint of original 1960 publication) UNESCO Ad Hoc Expert Group on Endangered Languages (2003) Language Vitality and Endangerment. Document submitted to the International Expert Meeting on UNESCO Programme Safeguarding of Endangered Languages Paris, 10–12 March 2003 (accessed 1 May 2010 at: http://www.unesco.org/culture/ich/doc/src/00120-EN.pdf) Woodward, J. (1994) Describing Variation in American Sign Language: Implicational Lects on the Deaf Diglossic Continuum. Burtonsville, MD: Linstok Press. (publication of author’s 1973 Georgetown University doctoral dissertation) Zeshan, U. (2006) Interrogative and Negative Constructions in Sign Languages. Nijmegen: Ishara Press. Zeshan, U. & Perniss, P. (2008) Possessive and existential constructions in sign languages. Nijmegen: Ishara Press.