Variations on polysynthesis: The Eskaleut languages (review)

This book is, to my knowledge, a first in English, bringing together research on aspects of Eskaleut languages (also called Eskimo-Aleut). These languages are or were spoken in the Arctic— Siberia, the Aleutian Islands, Alaska, Canada, and Greenland. They have been the topic of study by a fine group of scholars in recent years, and this book demonstrates the stimulating research that is taking place.

Perhaps Eskimo (as opposed to Aleut) languages are best known for their polysynthesis, or, in informal terms, their high number of morphemes per word. De Reuse (1994), following Anthony Woodbury, proposes the picture of the Eskimo verb in 1.

1. 1. base + postbaseⁿ_o + ending + enclitics^m_o

The base is the lexical core; postbases are derivational suffixes, and the endings mark inflection.

The structure in 1 alone does not immediately persuade the reader that these languages are polysynthetic. As the authors note, their polysynthetic nature becomes evident through multiple occurrences of postbases and enclitics. Some examples that illustrate the complexity of the word are given in 2.¹

1. 2.

   1. a. Central Siberian Yupik (de Reuse 1994:25)

      angyaghllangyugtuqlu

      angyagh-ghllag-nge-yug-tuq=llu

      boat-big.N-acquire.N-want.to.V-ind(3sg)=also

      ‘also, he wants to acquire a big boat’

   2. b. Central Alaskan Yupik (Ch. 6, p. 83)

      qaya-pi-li-sciigal-ngu-u-lria

      kayak-genuine-make-cannot-denom-be-rel.abs.sg.

      ‘the kind/type of one who cannot make a genuine canoe (the one who is onewho cannot make a genuine canoe)’

These words are ‘long’ in containing many morphemes. Mithun 1999 argues that they are words according to criteria that include native-speaker judgments and phonological characteristics. Eskimo languages thus are traditionally considered excellent exemplars of polysynthesis, and the title Variations on polysynthesis is apt for a book on languages of this group.

While the title might suggest that the papers all concern polysynthesis, the book is quite varied in its coverage. As the editors say in the introduction, its primary goal is ‘to situate the Eskaleut languages typologically in general linguistic terms, particularly with regard to polysynthesis’ (vii). The papers range from those that deal directly with polysynthesis to those that examine syntactic and semantic properties of the languages, to those that focus on discourse structure, and, finally, to those that examine language contact and change. The chapters in this volume present studies that center on the issue of what it means to be polysynthetic: they discuss synchrony and diachrony, including ongoing change; they ask whether inflection and derivation present sufficient categories to account for the rich morphology of Eskaleut and other languages; they question whether the verb is a word from a morphological perspective or is perhaps a word simply from a phonological perspective. Furthermore, the contributors raise questions of importance to linguistic theory, highlighting different theoretical perspectives, united by the languages of study. Many of the papers take up comparative discussion, within the family and beyond, in which languages as diverse as Chukchi, Mohawk, Apache, and Nu-Cha-Nuulth figure prominently. The editors sum up the volume by noting that it ‘shows how polysynthesis increases or decreases according to factors of internal development … or under the pressure of innovation resulting from bilingualism, language contact, or new forms of expression in the younger generation’ (viii). In the following discussion, I concentrate on the contributions to polysynthesis, with brief summaries of other topics.

As befitting a book with the word ‘polysynthesis’ in the title, Part 1 begins with articles that directly address the topic of polysynthesis—a topic of theoretical concern in recent years. In the Eskaleut tradition, polysynthesis is defined as above: languages with a high number of morphemes per word. Baker 1996, an influential work on polysynthesis, provides a narrower definition, one that excludes Eskaleut languages, since they lack structures that Baker views as critical for polysynthesis. In particular, Baker argues that polysynthetic languages use pronominal affixes and incorporated...