|AUTHOR: Wiklund, Anna-Lena
TITLE: The Syntax of Tenselessness
SUBTITLE: Tense/Mood/Aspect-agreeing Infinitivals
SERIES: Studies in Generative Grammar 92
PUBLISHER: Mouton de Gruyter
Cynthia L. Zocca, Department of Linguistics, University of Connecticut
The main claim of the book is that the three structures in (1), from (spoken)
Swedish, are surface variants of Tense/Mood/Aspect-agreeing infinitivals, in the
sense that they involve complementation and semantic vacuous tense/mood/aspect
inflection of the embedded verb. In other words, these embedded verbs are in
fact tenseless. Wiklund claims that the inflection of the embedded verb surfaces
through an Agree-type dependency between functional heads of the same label,
i.e. the embedded functional heads copy values from the relevant functional
heads in the matrix clause.
(1) a. Han fšrsškte o skrev ett brev --- Tense/Mood/Aspect-copying construction
he try.PAST & write.PAST a letter
'He tried to write a letter.'
b. Han hade kunnat skrivit. --- Participle-copying construction (PCC)
he had can.PPC write.PPC
'He had been able to write.'
c. Han satt o skrev dikter. --- Pseudocoordination (PC)
he sit.PAST & write.PAST poem.PL
'He was writing poems (in a sitting position).'
Based on the main claim, Wiklund proposes that there are three ways of being
tenseless: the relevant Tense (T) domain can be missing; the T domain can be
externally valued by the matrix domain (tense restructuring); or the T domain is
internally valued (no tense restructuring).
In order to investigate the nature of the structures in (1), Wiklund has to
delve into issues like restructuring, structure of CP (complementizer phrase)
and TP (tense phrase), complementation, and verb selection, making this book a
valuable source of data and discussion bearing on these topics.
In Chapter 1, Wiklund briefly introduces the structures considered in the book
and presents an overview of the book. TMACC verbs can alternate between a copied
form or an infinitive and the two verbs sharing their inflection are separated
by the element o(ch), which is homophonous to a variant of the conjunction
meaning 'and'. Also, copying can occur with the present, past, or imperative.
PCC verbs can only be in the past participle (PPC) and o(ch) is not allowed.
Finally, PCs cannot alternate with infinitival forms and, as in the case of
TMACC, copying is not restricted to the PPC. Also, the o(ch) element is present.
What the three constructions have in common is first of all that the two verbs
involved in each one have identical morphology, which is semantically vacuous in
the embedded one; also, only one subject, the matrix one, is licensed; finally,
the class of matrix verbs that can occur is restricted.
Chapter 2 deals with some apparent surface differences between TMACC and PCC,
showing how these two constructions are in fact similar: both of them involve
complementation, have semantically vacuous morphology in the embedded verb, and
involve copying that is syntactic, top-down, and local. What distinguishes TMACC
from PCC is how much functional structure can be copied: the full range of
verbal forms in the former but only the participial form in the latter. Another
difference is that the element o(ch) can occur between the two verbs in TMACC,
but not in PCC. Because both structures involve a copied element that has
semantically vacuous inflection, Wiklund calls them both ''copying infinitivals''.
In Chapter 3 we learn that copying infinitivals behave in much the same way as
standard infinitivals. This observation comes after an extensive survey of the
behavior of infinitivals in Swedish, dividing them into different classes based
on issues such as propositionality, factivity, infinitival markers, raising,
ECM, and control verbs. The author then presents the generalization that copying
is restricted to tenseless infinitivals, by which she means forms with no tense
inflection or with vacuous tense inflection. Furthermore, within the classes of
verbs that select copying complements, TMACC or PCC infinitivals are in
complementary distribution: the former are related to non-bare infinitivals and
the latter, with bare ones.
Chapter 4 draws on the relations between copying and non-copying infinitivals
from Chapter 3 and takes a closer look at the conjunction-like o(ch) introducing
a TMACC and the word att that introduces a standard infinitival, concluding that
they are both complementizers. This means that the matrix verb selects the same
category, CP, in TMA-copying and non-copying infinitivals. PCC infinitivals, on
the other hand, are, like bare infinitivals, a category smaller than CP. The
structure of copying and non- copying infinitivals that she arrives at are in
(2) and (3).
(2) Tenseless non-bare infinitivals (non-copying and copying):
[CP... [TP... [AspP... [vP... ]]]]
(3) Tenseless bare infinitivals (non-copying and copying):
[AspP... [vP... ]]
At this point Wiklund presents her proposal regarding the relation between TMACC
and PCC, namely that copying is proportional to the number of functional
projections in the matrix and embedded verbs. Copying, then, is a reflex of
dependencies between functional heads of the same label. In the rest of Chapter
4, she presents arguments in favor of taking copying to be a surface reflex of
restructuring. If this is correct, restructuring would not then be restricted to
Chapters 5 and 6 present more details about pseudocoordinations, showing that
they do not involve coordination, adjunct, or complex heads. Because they
contain semantically vacuous inflection in the embedded clause and a top-down,
local, and sensitive-to-tense relation between the matrix and embedded verbs,
Wiklund concludes in Chapter 5 that they are a case of TMACC. The difference
between TMACC and PCs is based on the restricted class of matrix verbs that can
license PC because of their aspectual properties. Chapter 6 looks at some
semantic properties of pseudocoordinating verbs, concluding that they
instantiate light verb uses of otherwise lexical verbs. Being light verbs, they
trigger restructuring of the event structure of the sentence.
In Chapter 7 Wiklund proposes an approach to derive the results of the previous
chapters. She takes the dependency involved in copying to be Agree, based on the
following observations about copying: c- command requirement, locality
constraints, and some kind of feature sharing. However, she concedes that it
must be a different kind of Agree, triggered by the lower verb. She also argues
for the need to reconsider approaches that explain restructuring in terms of
Finally, in Chapter 8 she concludes by suggesting a typology of tenseless
infinitivals based on the presence or absence of the T domain and on how T is
valued, internally or externally (i.e. by the matrix T) when it is present.
The book also has four Appendices. The first one is dedicated to less clear-cut
cases, involving Swedish verbs corresponding to 'dare', 'manage', 'succeed',
'pretend', and some psych predicates. Appendix II brings an overview of the
distribution of copying infinitivals in other Scandinavian languages, namely
Alvdalsmalet-Swedish, Norwegian, Danish, Faroese, and Icelandic. Appendix III
shows some selectional restrictions involving inanimate subjects. Finally,
Appendix IV points out that sometimes Agree results in a feature appearing
phonologically twice or sometimes only once.
The book is an invaluable resource for syntacticians, especially those
interested in working on Swedish, Scandinavian languages, and Germanic syntax in
general. The breakdown of the data is very clear, making the book convenient to
In the introductory chapter, Wiklund characterizes the book as offering ''a
detailed case study of agreeing infinitivals in Swedish'' (p. 3). As a case
study, the book provides a thorough and valuable portrayal of the relations
between the infinitival copying constructions in question. The arguments Wiklund
presents in favor of considering the three constructions as surface
instantiations of the same structure are indeed convincing.
In spite of the excellent presentation of data, the book does not do such a good
job in its more technical aspects, especially in Chapters 4 and 7. The first
problem is that since some TMA infinitives can also lack the infinitival marker,
that property cannot be used as evidence for or against the presence of C or T.
Therefore, the argument for the structures in (2) and (3) becomes a bit circular
and not so well motivated.
In chapter 7, the author takes the relation between the verbs in the matrix and
embedded clauses to be Agree in the sense of Chomsky (2000, 2001) based on three
similarities between copying and Agree: they both operate under c-command, they
are both subject to locality, and they both involve some kind of feature
sharing. She claims that an unvalued CFin (finiteness head in a split CP system
as in Rizzi 1997) in the embedded clause triggers Agree with the matrix finite
CFin. Copying then is dependent on a finiteness head that is itself unvalued (or
missing, as in non-CP infinitivals).The first problem with this analysis is that
Agree is phase-bound, a phase being standardly taken to be vP or CP. Because in
Wiklund's analysis T-copying can apply across a CP, this shows that her version
of Agree is not subject to standard locality. Another problem, which Wiklund
herself acknowledges, is that whereas Agree is bottom-up (in the sense that the
value is copied from a lower goal to a higher probe), copying is top-down, since
tense valuation is copied from the matrix to the embedded verb. She calls this
kind of Agree ''iAgree'', where ''i'' stands for inverse. Considering it to be
outside the scope of the book, Wiklund does not discuss further implications of
a system that allows for an iAgree.
The richness of data in the book can also raise other issues, for example the
difference between Agree and agreement. It is hard to argue that all instances
of agreement do involve Agree, as for example in the case of DP-internal
agreement (also known as Concord). In the case of copying infinitivals, it is
quite clear that there is some agreement relation involved, but maybe it does
not necessarily involve Agree. This of course is really beyond the sope of the
book, but it would be interesting to see some future analysis of Wiklund's data
from a more morphological perspective, maybe involving PF-copying or late insertion.
Chomsky, Noam. 2000. Minimalist Inquiries: The framework. In R. Martin et al.
(eds.), Step by Step: Essays in Minimalist syntax in honor of Howard Lasnik, pp.
89-155. Cambridge, Mass.: MIT Press.
Chomsky, Noam. 2001. Derivation by Phase. In M. Kenstowicz (ed.), Ken Hale: A
life in language, pp. 1- 52. Cambridge, Mass.: MIT Press.
Rizzi, Luigi. 1997. The Fine Structure of the Left Periphery. In L. Haegeman
(ed.), Elements of Grammar: A handbook of Generative Syntax, pp. 281-337.
ABOUT THE REVIEWER
Cynthia L. Zocca is a doctoral student at the University of Connecticut. Her
research interests include syntax, semantics, and morphosyntax, especially of
Brazilian Portuguese and its relations to other languages.