Christensen, Birthe H. 2001 Dialekten i Sydvestgrønland: en grønlandsk i-dialekt. Eskimologis Skrifter, nr. 16, Københavns Universitet, 87 pages [The Dialect in South West Greenland: A Greenlandic i-dialect. Papers from the Department of Eskimology no. 16, University of Copenhagen].

The publication is — apart from minor biographical adjustments — identical to a prize paper at the University of Copenhagen. Birthe H. Christensen (BHC) has investigated the Southwest Greenlandic dialect (as spoken in Narsaq and Qaqortoq municipalities), a so-called "i-dialect." The term i-dialect refers to a characteristic phenomenon common to certain Greenlandic dialects where / u / is replaced by / i / according to more or less complicated rules. The purpose of BHC’s investigation is manifold: to describe a hitherto not described dialect and to compare it to the neighbouring dialect of Kap Farvel (described by Rischel 1974, 1975); to trace phonological and morphophonological changes through a comparison of speakers of different ages (the oldest born in the mid-1880s, the youngest born in 1969), in order to see if the dialect is approaching standard language (i.e. Central West Greenlandic). BHC expects so, because of the increasing impact of mass media and because of internal migration in Greenland. Interestingly, BHC also suggests that the very complexity of the rules might lead to a change which will be reflected, first and foremost, in a greater degree of lexicalized forms. Finally BHC discusses the possible origin of the i-dialect phenomenon through comparison with other Eskimo languages.

The book is divided into five parts: 1) background, 2) transcriptions (in broad phonemic notation) of 12 recordings, 3) analysis of the data material and a conclusion, 4) references and 5) appendices (glossed text excerpts).

In the first part, BHC gives a short presentation of previous descriptions of the i-dialect phenomenon. In the early sources, the i-dialect phenomenon was seen as a more or less arbitrarily occurring variation, until Rischel demonstrated in 1974 and 1975 the regularities behind the i-dialect phenomenon. The rules can only be formulated as "rules for / u / not changing into / i / ." Those are the so-called "Rischel’s Laws" (Dorais 1981): never in the first syllable (e.g., suli), never after a labial (e.g., immuk), and never after a preceding / u / (e.g., immussuaq). Furthermore there is a tendency that a following / u / also protects against changing (e.g., irnusuttuq), except when there is an intervening labial consonant (e.g., irnisippuq). A few years later, a comprehensive hypothesis concerning the spreading of the i-dialects was put forward (Fortescue 1986; Petersen 1986; Rischel 1986). In the first part of her book, BHC also presents the two different theories of the origin of the i-dialect phenomenon, namely diphthongization (Rischel 1974, 1975) and delayed labialization (Fortescue 1984).

The data material of BHC consists of both sentence lists and interviews, partly her own recordings, party recordings (tape and video) by others. The purpose of using sentence lists is to make sure that the relevant structures are present. BHC mentions the methodological problems in using sentence reading and interviews: the style is more or less formal, and perhaps the speaker speaks a more standard-like variety, both due to the more formal situation in itself, and for reasons of being "polite" to the interviewer. This was the case of at least one speaker whom BHC easily communicated with during the interview, but whom BHC hardly understood while she was speaking to her mother on the telephone.

The second part of the book consists of a (broadly phonemic) transcription of the recordings, which, together with the glossed transcriptions of part five, is a valuable documentation of the dialect.

The third part is the main part where BHC goes into great detail in her description and analysis of the recordings. BHC focuses on specific features, e.g., what happens when / u / occurs in two adjacent syllables? In Kap Farvel there is a "backwards" protection rule, which is blocked by an intervening labial consonant (not irnisuppuq but irnisippuq). BHC finds evidence (according to her expectations) that this complicated rule is beginning to disintegrate among the younger speakers, partly due to an increasing degree of lexicalization, partly due to individual variation. BHC even finds that one of the speakers seems to find alternative words so that she avoids taking the rule into consideration.

But BHC has another, more surprising result: in the sentence reading list, the youngest speaker has relatively more lexicalizations than she has in the interview. This means that the complicated rule is utilized almost without exception in the spontaneous speech, but vacillating in the controlled, formal speech situation. From the point of view of language change and language attitudes, this is very interesting and might have deserved more attention.

BHC is a not only a very careful observer, she is also very cautious not to draw untenable conclusions. For example, BHC is very concerned about how to decide whether an unexpected form is the result of vacillating rules or is a result of lexicalization. For example, when the youngest speaker used -imaar in nuukkimaarluta ["that we wanted to move"] instead of the expected ‑umaar ["want to"] (protected by the preceding / u / of the stem), BHC suggests that this is a lexicalized form, because this speaker elsewhere says -imaar with / i /, regardless of the phonological context. Yet, when BHC wants to investigate whether an intervening labial blocks the regressive assimilation for the oldest speaker, she concludes: "[…] three times that / u / is preserved on both sides of a labial consonant; but there is no possibility of comparison with other derivations of the same word" (p. 55) [my translation]. Hence, nothing can be concluded as regards lexicalization for this speaker.

One of the most important results of BHC’s investigation is that the / u / - / i / rules are still productive even among the youngest speakers, although there is individual variation and perhaps an increasing degree of lexicalization. The pattern of Southwest Greenlandic is similar to that of Kap Farvel, except that the most complex rule, the blocking of regressive assimilation, is declining.

In the last paragraphs of the book BHC returns to the theoretical question of diphthongization vs. delayed labialization. Turning to other Eskimo dialects, BHC takes both synchronic and diachronic evidence into consideration, and tries to reach a synthesis: the / u / to / i / change may have started as a diphthongization ( / u / > [iu]) and the labial element of the diphthong may have been delayed / moved to the next segment where it could protect the labiality of adjacent / u / ’s.

In her book, BHC gives a valuable contribution to the description and understanding of an interesting dialectal phenomenon. In her quasi-longitudinal investigation, she sheds light over some possible mechanisms behind sound change and diffusion, not only through time and through speakers, but also through lexicon. Her main focus is phonetics, but as she also takes pragmatic and semantic factors into consideration — as well as comparative — her book will be of interest to many readers.