Distributed Morphology Today is many things. The stated purpose of this volume is to be a Festschrift to celebrate the 90th birthday of Morris Halle. To this end, the volume is composed of twelve short papers from Halle's former students, several of which have taken the time in the endnotes to say very nice things about him. Indeed, in particular, the introductory chapter and Alec Marantz's chapter incorporate very pleasant and nostalgic blurbs about the authors' experiences working with Halle. Overall the volume has a distinct feeling that it was put together by its contributors to express gratitude, love, and admiration for their mentor/teacher/colleague, which is exactly, to my mind, how a Festschrift ought to feel.
However, Halle's long, extremely important career in morphological theory is not limited to his being one of the founders of the theory of Distributed Morphology (DM). That this Festschrift chooses to focus on DM is still fitting, though, because its publication nicely aligns with the 20th anniversary of the original proposal of DM. As a result, another role this volume plays is to serve as a ‘state-of-the-theory’ volume for DM. In this role, it offers an apt snapshot of current research in DM. Most of the articles fit a fairly typical rubric found in research of DM in that they tackle a data set and then use the explanation of that data set to argue for or against a modification to the architecture of DM. Those articles that don't as nicely fit the typical DM rubric, e.g. Matushansky's and Marantz's chapters, also feel like they belong in a ‘state-of the-theory’ volume. The former presents data that has already been accounted for twice within DM, discusses those analyses within the confines of the theory, and argues for a new, preferable account within DM. The latter brings DM into a tragically understudied part of the architecture of the grammar within this framework: the interface with semantics. Indeed, this volume so much represents ‘state-of-the-theory’ that the reference section reads like a reading list for who is working in DM and what they have recently been working on.
Finally, one of the great strengths of DM (and item-and-arrangement, morpheme-based models of morphology more generally) is that it accounts for locality effects quite nicely. Because of this, it is not surprising that a persistent theme among the articles in this volume is locality effects. Indeed, outside of its function as a Festschrift and a ‘state-of-the-theory’ volume, this volume clearly also might be taken to be a volume on locality effects in morphology. It is notable, however, that quite a few of the locality effects discussed are triggered by a null morpheme, by theorized syntactic heads, and/or by phase boundaries (Embick's chapter is a particularly good example). This has the noticeable effect of making the overall volume read not as a general volume on locality, but more on the locality that is observable if you happen to accept nearly all the underlying assumptions of DM, the Minimalist Program, and Phase Theory. For that reason, morphologists who are not practitioners of DM will likely not find the arguments to be very compelling.
Chapter 1 of Distributed Morphology Today is ‘Variability and allomorphy in the morphosyntax of Catalan past perfective’, by Isabel Oltra-Massuet. The central concern of this chapter is three different forms of the Catalan past perfective: a synthetic form (purificares), an analytic form (vas purificar), and a non-standard hybrid (vares purificar) (all forms are ‘purify.2sg.pst.perf’). The analysis provided by the author makes for a very good chapter for this volume as it employs much of the architecture of DM. For example, Oltra-Massuet employs an impoverishment rule, dissociated morphemes, fusion, and contextual licensing of vocabulary insertion. Her chapter is also one of the set that forms a small sub-theme of the papers here that deal with the topic of root suppletion in some way.
Chapter 2, ‘Phonological and morphological interaction in Proto-Indo-European accentuation’, is Rolf Noyer's contribution and is a noticeable exception to many of the generalizations I made above. This chapter makes almost no reference to the architecture of DM at all. Rather, it is a return to Halle's (Reference Halle1997) preliminary analysis of (the Schindler-Rix reconstruction of) PIE accentuation within the architecture of Simplified Bracketed Grid Theory (Idsardi Reference Idsardi1992). Noyer expands upon and further develops Halle (Reference Halle1997) with the primary focus on generating the reconstructed accent classes and only those classes.
Martha McGinnis's Chapter 3, ‘Agree and fission in Georgian plurals’, is primarily concerned with the universal hierarchy of features (Harley & Ritter Reference Harley and Ritter2002) and the nature of feature discharge. The primary data under discussion here is the apparent double marking of plural in Georgian number agreement. McGinnis argues that what appears to be a second reflex of plural is in fact a realization of a feature she proposes: multi-speaker (a dependent of speaker, participant, and person). Perhaps the most novel of her proposals is one involving a class of features that self-delete if they are not discharged rather than causing the derivation to crash.
As I alluded to above, Ora Matushansky's Chapter 4, ‘More or better: On the derivation of synthetic comparatives and superlatives in English’, is a return to well known (and persistently problematic) alternation between -er and more (and -est and most) in English. Matushansky argues against the two lowering proposals extant in DM (Embick & Noyer Reference Embick and Noyer1999 et seq. and Bobaljik Reference Bobaljik2012) and for a head-movement analysis that is supplemented with ‘much-support’.
Chapter 5, ‘Is word structure relevant for stress assignment?’, by Tatjana Marvin, is a comparison of two accounts of the familiarly tricky English stress: DM and Optimality Theory (OT). Marvin proposes a DM analysis of English derivation that shows that phase-based cyclic spell-out accounts for why some affixes like the famous -able sometimes trigger stress shift and sometimes do not. Marvin argues that -able can attach both inside and outside a phase head, predicting this behavior. She compares her analysis to Burzio's (Reference Burzio1994) analysis and concludes that the phase-based analysis is superior because such an account is not available to (non-Stratal) OT.
Marantz's contribution, Chapter 6, ‘Locality domains for contextual allomorphy across the interfaces’, is a special paper in this volume and as such is honestly quite a bit of fun to read. In the paper, Marantz spends a lot of time discussing the development of the theory of Distributed Morphology in a casual style. The effect is that it feels like the reader is being told stories over dinner about where DM comes from. As such, its presence in the book is wonderfully welcome. The content of the chapter is that the well-studied interface with the phonology has an analogue with the semantics interface: just as there is the familiar contextual allomorphy (a root takes on different forms in different environments), so too is there contextual allosemy (a root takes on different meanings in different environments) and both are constrained in the same ways by locality constraints that derive from phase boundaries.
Heidi Harley & Mercedes Tubino Blanco's Chapter 7, ‘Cycles, vocabulary items, and stem forms in Hiaki’, is primarily an analysis of the class-based context-driven allomorphy in Hiaki stems. However, it also contributes two key architectural arguments. First, the authors argue that class features are not syntactic features (contra Embick & Halle Reference Embick, Halle, Geerts and Jacobs2005), but are rather properties of the inserted Vocabulary items (which then license readjustment rules). The second key argument found here is more evidence in support of a readjustment approach to stem allomorphy rather than a listing approach (which is assumed by the majority of the relevant papers in this volume).
Daniel Harbour's Chapter 8, ‘“Not plus” isn't “Not there”: Bivalence in person, number, and gender’, is primarily a survey chapter that presents quite a bit of evidence strongly suggesting that features are bivalent and not privative. The evidence chiefly comes in two classes: (i) that negation of a feature and absence of a privative are fundamentally different; and (ii) that morphological operations have to make explicit reference to negative features.
David Embick's Chapter 9, ‘Morphemes and morphophonological loci’, presents certainly the biggest picture analysis of the papers. The goal of this paper is to add another argument in support of morpheme-based approaches to morphology (as opposed to affixless or word-based models). To this end, Embick examines several apparently nonaffixal alternations and argues that these effects are indeed conditioned by the internal morphosyntactic complexity in the words and, moreover, the effects are restricted by locality. He proposes the ‘Morphological Locus Theorum’ (153), which asserts that triggered morphological changes must be local to the triggering head.
Like Chapter 2, Chapter 10, ‘Agreement in two steps (at least)’, by Eulàlia Bonet, is a bit of an outlier in that it does not propose any revisions to the architecture of DM nor does it argue for DM. Rather, Chapter 10, in providing an account for nominal agreement systems, assumes an interface between DM and OT and focuses primarily on an OT account which proposes that agreement is the result of two steps, one syntactic (employing DM) and one post-syntactic (employing OT).
Chapter 11, ‘Suspension across domains’, by Jonathan Bobaljik & Susi Wurmbrand, bucks the locality theme of the volume by showing phenomena that seem to violate strict locality (e,g. suppletion in comparatives and superlatives, QR, and control/raising). Bobaljik & Wurmbrand argue that each of these cases involves what would typically be considered a phase head but in these cases do not function as phase boundaries. These phase heads are ‘deficient’ because they are dependent on an element higher in the tree (for example, subjunctive mood must be licensed by the superordinate clause). This deficiency gives rise to seeming penetrability of some phases and non-locality effects.
The final chapter, ‘Contextual neutralization and the Elsewhere Principle’, by Karlos Arregi & Andrew Nevins, argues for a subtle but radical change to the fundamental operation of DM: Vocabulary insertion. Insertion, as Arregi & Nevins point out, is constrained in three ways: underspecification of features, the Elsewhere Principle, and licensing into specific environments. These three are importantly ranked – morphosyntactic features are more important to insertion than context. This chapter argues, on the basis of several examples of context neutralization, that rather contextual specificity outranks the Elsewhere Principle. This means that in some cases contextual licensing is more important than morphosyntactic features.
Overall, I was quite happy to read this volume. As a Festschrift, it was enjoyable to read in a way that most edited volumes containing only loosely related papers are not: it inspired the persistent feeling of nostalgia and respect for Halle's work. I also think that this volume is very representative of currently ongoing DM research. Indeed, this volume presents typical work of the most prominent practitioners of DM. This is also, though, something I perceived as a bit of a weakness of the volume. As someone who is familiar with the DM literature, almost every chapter felt very familiar. In many cases, the papers were short expansions of previous work, so while the arguments put forth here are original, they importantly did not feel novel.
Since DM itself is a framework upon which many different theories are built, it is to be expected that this volume is not internally consistent, but it is still striking that the chapters are often not compatible with each other, sometimes on fundamental assumptions. For example, some papers crucially assume Vocabulary insertion is after linearization (Embick and Arregi & Nevins) while others (those that make reference to hierarchal structure at insertion; most of the rest of the volume) assume it is before linearization. As mentioned above, some papers require that root Vocabulary Items compete with each other while others require a readjustment approach. In addition, Bonet assumes an OT interface with the phonology while the other authors rely on a rule-based phonology.
Paradoxically, this volume also seems much more homogenous than research in DM really is. A great many of the more radical architectural proposals for DM are not represented here. For example, noticeably absent are representative models which limit DM's morphological component to insertion only, such as Trommer (Reference Trommer, Retoré and Stabler1999), or models where insertion is not limited to terminal nodes, such as Radkevich (Reference Radkevich2010) or Svenonius (Reference Svenonius2011). This is probably a result of the Festschrift function of the volume. Perhaps the more radical departures from the architecture of DM simply are not especially common among Halle's former students, but, on the other hand, these other approaches would happily be included in a volume covering ‘DM today’ in its fullest scope.
