Zeller, Jochen (2001) Particle Verbs and Local Domains.
John Benjamins Publishing Company, xi+323pp, hardback ISBN
1-58811-036-2 (US), 90-272-2762-4 (Eur), USD 99.00, EUR 109.00,
Linguistik Aktuel/Linguistics Today 41.
Peter Svenonius, University of Tromsoe
Scores of scholars have been tempted into the lush jungles of
the separable-prefix or verb-particle construction, never to
return. But later expeditions have been better equipped and
better prepared, and gradually, have hacked paths deep into
the jungle, making maps (of varying quality) and reaping
theoretical fruits and gems from the dark and tangled
The work reviewed here (essentially the author's 1999
University of Frankfurt dissertation) represents a well-
prepared expedition with the newest equipment (though not all
of the old maps). It defends a novel analysis of the
construction, based chiefly on German data. This book makes
an important contribution, especially with regard to the
syntax of the verb-particle construction.
Here I summarize the contents first and then discuss some of
the findings. The analysis is stated over German (and Dutch),
but is briefly extended to English and Mainland Scandinavian
in the final chapter. Thus I use English examples here in my
discussion in some cases to illustrate points which Zeller
makes with German ones. When I repeat Zeller's German
examples, I occasionally make use of more literal word-by-
word glosses than he does.
Chapter 1: Syntax, morphology, and lexical licensing
Chapter 2: The syntax of particle verbs
Chapter 3: The non-functional nature of particle phrases
Chapter 4: The lexical representation of particle verbs
Chapter 5: Local domains and morphology
Chapter 6: Particle verbs and word formation
Chapter 7: Typological remarks and reanalysis
DESCRIPTION OF THE CONTENTS
For readers who might be unfamiliar with the construction
which is the concern of this book, I illustrate it in (1-2),
a minimal pair with the separable particle verb _umfahren_,
'run over' in (1), and the inseparable prefix verb _umfahren_
'drive around' in (2); both are written the same and
literally calque to "about-drive," but the prefix is
unstressed while the particle is stress-bearing (cf. Zeller's
examples on p. 57).
(1) a. ...weil Peter den Mann umfaehrt
because Peter the man about.drives
'...because Peter runs the man over'
b. Peter faehrt den Mann um.
Peter drives the man about
'Peter runs the man over'
(2) a. ...weil Peter den Mann umfaehrt
because Peter the man about.drives
'...because Peter drives around the man'
b. Peter umfaehrt den Mann.
Peter about.drives the man
'Peter drives around the man'
The (b) examples show the effect of verb movement, stranding
the particle in (1b) but carrying along the prefix in (2b).
The book examines the partly syntactic, partly morphological
nature of particle verbs, and argues that they are like
syntactic complementation constructions in involving phrasal
projections combined in the syntax, but that unlike the
situation with ordinary syntactic complementation, the
particle lacks functional structure. The notion of
'structural adjacency' is introduced, essentially a relation
holding between a head and the head of its complement.
An old debate is to what extent particle verbs are wordlike.
For example, they productively allow nominalization,
retaining the idiosyncratic peculiarities of the verbal
construction ('our playing down of the differences' means
what you would expect from the idiomatic 'play down,' meaning
'make to appear less significant'). Zeller gives some
additional purported wordlike properties in Chapter 5,
In some analyses, particle verbs are taken to be essentially
words, with the special property that they can be separated
by syntactic operations like verb movement. Zeller argues at
length against this position in Chapter 2, marshalling a
range of syntactic arguments that particles are phrasal
(generally PPs, but also in some cases APs or NPs), generally
building on the separability of the particle from the verb
(e.g. in 'Off he went,' cf. p. 89), but also on some other
features, for example the morphological independence of the
verb from the particle (in terms of inflectional classes) and
the possibility of modification of the particle (in examples
like 'turn the oven all the way up,' cf. p. 100).
In Chapter 3, Zeller argues that although particles are
phrasal (PP, AP, or NP), they are different from most phrases
in lacking a functional projection (DP for NP, AgrP for AP,
and FP for PP). This is a key argument which arises several
times later in the book to account for special properties of
the verb-particle construction, mainly because of the notion
of "structural adjacency," a relation which is defined in
such a way that functional structure disrupts it. Any
relation between the verb and the particle that is dependent
on structural adjacency cannot be maintained if a functional
Bare Ns in particle verbs like _Karten spielen_ 'play cards'
(literally 'cards play') are plausibly NPs, as opposed to
DPs. As Zeller points out, they do not introduce discourse
referents ('We played cards... #They were new'), and resist
modification (*'We played new cards'; cf. p. 130). But
motivating the absence of functional projections over the
other categories is less straightforward. Zeller takes
referentiality to be central in distinguishing lexical
projections from functional ones.
For AP, Zeller notes that resultatives show agreement in
languages with agreement on predicative adjectives (such as
Norwegian), and suggests that resultatives do contain an AgrP
dominating AP. He suggests that this is connected to
referentiality (p. 146), and proposes that non-resultative
particle verbs (such as _krankfeiern_, lit. 'sick.celebrate,'
meaning 'play hooky') involve bare APs.
Similarly, he contrasts the referentiality of prefixed
particles (such as _heraus_, lit. something like
'hither.out') with the non-referentiality of non-prefixed
particles to support the idea that the former are FPs
containing PPs, while the latter are bare PPs. He illustrates
this with the following examples (p. 139, citing McIntyre
(3) a. Peter will einen Kreis ausschneiden.
Peter wants a circle out.cut
'Peter wants to cut out a circle'
b. Peter will einen Kreis herausschneiden.
Peter wants a circle HER.out.cut
'Peter wants to cut a circle out (of some unspecified
As another argument for the lack of an FP layer over the
particle, Zeller introduces a proposal regarding case
assignment by prepositions, where the case assignment
properties of prepositions rely on the functional layer FP.
If particles lack this layer, they will be unable to assign
Chapters 2 and 3 constitute a well thought out and well
reasoned approach to the special status of the verb-particle
construction. The rest of the book treats their semantic and
morphological status, using advanced tools made available by
Jackendoff, Marantz, and Borer.
Essentially, Zeller's approach to the meanings of particle
verbs is that they are idioms (Chapter 4), though the whole
picture is somewhat more complex.
First, since there are productive patterns of particle verb
formation, there must be ways to systematically extend their
inventory. Here Zeller adopts Stiebels & Wunderlich's (1994)
proposal that specific meanings for particular particles can
be listed along with a characterization of the class of verbs
with which they occur. This approach is discussed further in
Chapter 5 with an extended analogy being drawn between the
context-sensitive semantic contribution of the particle and
context-sensitive morphological phenomena such as allomorphy
and stem suppletion (section 5.1).
Zeller also proposes a condition (p. 169) on particle verbs
which is stated in terms of the notion of structural
adjacency. It asserts that the special meaning of a particle
is only available when the particle is structurally adjacent
to V. This is intended to characterize the absence of the FP
projection, but also the absence of particles from non-
derived nominals (*_der Sprung auf_ 'the jump up,' p. 171).
The condition of structural adjacency is engineered to allow
the statement of a notion of "reanalysis," which is
introduced in Chapter 6 to allow morphological structures
based on particle verbs to surface as V-zero. The definition
of reanalysis (p. 255, p. 257) is constructed to derive
several results, for example to prevent a complex P-V from
moving to C in V2, while still allowing it to move to the
right of V in Dutch verb-raising structures (detailed in Ch.
Finally, in Chapter 7, the analogous constructions in Dutch
and the Scandinavian languages are (rather cursorily)
compared with the German one. Zeller assumes that particle
shift options (i.e. the options of 'throw the dog out' and
'throw out the dog') reflect the optionality of reanalysis:
if PP is a phrasal complement to V, then it is left behind
when V moves to v, across the direct object. If V-P is a
complex head, then P is carried along when V-P moves. This
makes Zeller's analysis of particle shift much like those of
Larson (1988) and Johnson (1991).
The book is a solid and careful piece of research, especially
in the first half. It is methodical and clearly written,
painstakingly explaining the motivations for assumptions,
detailing alternative proposals, and arguing that they should
be rejected. Some readers will feel that they did not need so
much convincing that particles are phrasal, being satisfied
with an example or two of topicalization. But since the
morphological analysis lives on, Zeller hammers the point
This book can definitely be used in advanced instruction. I
found chapters 1-3 so clearly written and well reasoned that
I assigned them to my class of twelve master's students to
read. The students agreed that Chapters 2 and 3 were clear,
but they seemed to lack the background to fully appreciate
the discussion of Jackendoff, Marantz, and Borer in Chapter
The two most distinctive objects which Zeller brings back out of
the jungle are his particular notions of 'structural adjacency'
and 'reanalysis' I critically examine them in turn (the first
by way of the question of the presence of functional structure),
and then summarize.
Problem 1, the lack of functional structure above the
particle (Chapter 3).
The idea that particles lack functional structure is
important and potentially explanatory in addressing the
particle's intermediate status between a bound morpheme and a
full-fledged phrase. It is therefore worth a careful
examination. Zeller's use of referentiality as a diagnostic
for functional structure seems problematic; the
'referentiality' inherent in (3b) surely comes from the
incorporated element _her_, which has some deictic function,
and on Zeller's assumptions this implies a layer of
functional structure. But the connection between the deixis
and the FP above PP is less than obvious.
Zeller (pp. 127-128) invokes Jackendoff's (1983) notion of a
representation in a mental 'projected world' in discussing
referentiality, but it does not seem that these notions are
sharp enough to show why 'in' is not 'referential' in "Come
in" (e.g. as a response to a knock on the door).
Furthermore, some of the arguments for the lack of functional
structure in particles seem compromised by the very arguments
given in Chapter 2 that they are phrasal; for example, in
Chapter 2, examples are given of particles being modified by
adverbial material, but in Chapter 3 the impossibility of
adjectives with nominal particles is claimed to be evidence
that they do not project DP (cf. the discussion above about
'playing (*new) cards').
I found the non-case-assigning properties of particles a more
pleasing consequence of their lacking functional structure,
though it is weakened slightly by the dative-taking particle
verbs discussed in Chapter 5, section 5.2.
Perhaps more worrying is the question of what exactly
determines when functional structure is projected. Zeller
seems to assume that in general, functional layers will be
present unless blocked. The mechanism blocking them in the
verb-particle construction is the lexical entry, in which the
special meaning of the P-V combination is stored. Structural
adjacency plays a special role in the storage of lexical
meanings (Chapter 5, section 5.2), and structural adjacency
is disrupted by functional structure.
However, given that idiomatic meanings can be stored for
idioms containing functional structure (e.g. 'kick the
bucket'), it is unclear what status Zeller's structural
adjacency constraint actually has. In fact, Zeller explicitly
weakens it when he allows (in 5.3) the complex particles with
_hin_ and _her_ to have special meanings, on the basis of the
structural adjacency of V and F, since the relationship is
actually between a V, an F, and a P (e.g. in _herumreiten
auf_, lit. 'HER.about.ride on,' meaning 'to harp on' (p.
In this context it is worth thinking about what the small
clause analysis has to say about particle constructions.
(Zeller does not discuss the small clause analysis in any
detail.) Given that the predicate is phrasal, the small
clause analysis actually requires a functional head. This is
because, if the argument were to be projected in the
specifier of the lexical projection of P, then that argument
would be carried along by movement, for example in the
topicalization examples ('Away we chased the dog,' not *'The
dog away we chased').
It also follows from reasonable assumptions that the
functional head in the small clause would select a lexical
predicate, or at least something other than the argumental
functional structure, e.g. it would select NP rather than DP,
and plausibly might also take PP rather than FP. As for the
adjectival phrase, it is possible that the small clause is a
type of AgrP. Now it seems that Zeller's observations about
the lexical nature of the particle might follow from the
properties of the small clause head rather than from his
proposed restriction on particle meanings (that they are
licensed only in the context of structural adjacency with V).
Problem 2, Reanalysis (Chapter 6, Chapter 7).
I am skeptical about Zeller's notion of reanalysis; it seems
that a productive rule of lexical compounding would stand a
better chance of solving the problems raised by
nominalizations and other derived forms.
Reanalysis, according to Zeller, allows the insertion of a V-
P in Scandinavian, to derive the V-P-Object order ('throw out
the dog'). But this type of analysis has never been able to
satisfactorily deal with the fact that P follows V, rather
than preceding it, since other complex words in English and
Scandinavian are right-headed (especially since Scandinavian
does have incorporated P-V structures, for example in
passives, as Zeller notes in Chapter 7).
Zeller must place several restrictions on reanalysis to
prevent it from overgenerating. For example, he states that a
reanalyzed P-V in German must be further derived (p. 257), in
order to prevent P-V from appearing in the second position in
V2 clauses, while still allowing very free nominalization and
adjectivalization of particle verbs.
It seems that these and other facts might follow from some
reasonable assumptions about compounding. For example, German
(like English) freely allows N-N compounding but not N-V
compounding. Thus is it possible that similar considerations
would allow free P-N compounding (and P-A compounding), but
not P-V compounding, replacing Zeller's condition that
reanalyzed P-V's be further derived. The meanings of the
compounds would derive from the independently needed lexical
entries for the particle verbs, which would rule out P
compounding with non-derived N.
Zeller is also forced to assume that there is a lexical entry
for 'paint green' in Dutch but not for 'paint violet', since
the former but not the latter allows Verb Raising (i.e. the
order '...because John the door will green-paint').
Possibly, considerations of the productivity of compounding
might explain why 'green paint' is better than 'violet
paint'), without actually requiring 'green paint' to have its
own idiosyncratic lexical entry. Interestingly, Snyder (2001)
has argued for a very direct connection between the
possibility of compounding and the possibility of
resultatives, particle verbs, and other complex
constructions. On the other hand, the considerations that
Zeller raises in Chapter 2 all suggest that Dutch _groen
verven_ 'green paint' is not a compound at all, or a
reanalyzed word, and that Verb Raising should be analyzed in
a different way.
On the whole, this book is a major contribution to our
understanding of verb-particle constructions and is a must-
read for anyone who is serious about the construction.
Some may fault this work for its treatment of previous
literature on verb-particle constructions, which could have
been more thorough, but I feel that it makes up for any such
shortcomings with its originality and the careful attention
it pays to the need for precision. Actually, this very
strength may also be its greatest drawback, since it is at
times almost maddeningly explicit, for example, the
discussion on pp. 163-165 of the fact that V-to-C does not
interfere with idiomatic readings of V and its complements,
where many authors would have been satisfied with a brief
footnote saying that verb movement does not generally seem to
affect semantic interpretation. The thought occasionally
crosses a reader's mind that the book could have been a
little shorter -- but only occasionally.
I end with a few comments about the physical qualities of the
It is handsomely bound and professionally typeset, in keeping
with the usual standards for Benjamins' Linguistics Today
series (perhaps I should mention as a disclaimer that I have
edited a volume in this same series). The index is useful,
there are convenient footnotes rather than endnotes, and I
noticed very few typographical or printing errors.
Borer, Hagit. 1991. The causative-inchoative alternation: A
case study in parallel morphology. The Linguistic Review
Jackendoff, Ray. 1983. Semantics and Cognition. Cambridge,
Ma.: MIT Press.
Johnson, Kyle. 1991. Object positions. Natural Language and
Linguistic Theory 9:577-636.
Larson, Richard. 1988. On the double object construction.
Linguistic Inquiry 19:335-391.
McIntyre, Andrew. 2001. German Double Particles as Preverbs:
Morphology and Conceptual Semantics. Tuebingen: Stauffenberg
Snyder, William. 2001. On the nature of syntactic variation:
Evidence from complex predicates and complex word-formation.
Stiebels, Barbara, & Dieter Wunderlich. 1994. Morphology
feeds syntax: The case of particle verbs. Linguistics 32:913-
ABOUT THE REVIEWER
Peter Svenonius did his undergraduate work at the University of
Massachusetts at Amherst, and received a PhD in linguistics from
the University of California at Santa Cruz in 1994. He has been
teaching ever since at the University of Tromsoe, which is much
warmer than you would expect for a place two hundred miles north
of the Arctic Circle, but just as spectacular.