Incorporating "Unconscious Reanalysis" into an Incremental, 
Monotonic Parser 
Patrick Sturt* 
Centre for Cognitive Science 
Edinburgh 
UK 
sturt@cogsci.ed.ac.uk 
Abstract 
This paper describes the author's imple- 
mentation of a parser aimed at repro- 
ducing, in a computationally explicit sys- 
tem, the constraints of a particular psy- 
cholinguistic model (Gorrell in press). In 
Gorrell's model, "unconscious" garden 
paths may be processed via the addition 
of structural relations to a monotone in- 
creasing set at the point of disambigua- 
tion, but there is no discussion as to how 
the parser decides which relations to add. 
We model this decision as a search for a 
node in the tree at which an explicitly 
defined parsing operation, tree-lowering 
may be applied. With reference to En- 
glish and Japanese processing data, we 
show the importance of this search for 
empirical adequacy of the psycholinguis- 
tic model. 
1 Conscious and Unconscious 
Garden Paths 
Certain researchers in the psycholinguistic com- 
munity (Pritchett (1992), Gorrell (in press)), have 
argued for a binary distinction between two dis- 
tinct types of garden path sentences. Conscious 
garden paths, such as (1) below, are locally am- 
biguous sentences which give rise to reanalysis 
that is both experimentally detectable and causes 
a conscious sensation of difficulty or "surprise ef- 
fect". Unconscious garden paths, on the other 
hand, such as (2), cause reanalysis which is exper- 
imentally detectable, but which is generally not 
"noticed" by the speaker or hearer. 
*The work reported here was done very much in a 
collaborative spirit with my supervisor, Dr. Matthew 
Crocker, and thanks are due to him for innumerable 
suggestions and ideas. I would also like to thank the 
people who have offered insightful comments on this 
work, in particular, David Milward and Martin Pick- 
ering. The research was supported by ESRC grant 
R00429334338 
(1) While John was eating the ice cream melted. 
(2) John knows the truth hurts. 
This binary distinction has often been used to mo- 
tivate a two-level architecture in the human syn- 
tactic processing system, where what we will call 
the "core parser" performs standard attachment, 
as well as being able to reanalyse in the easy cases 
(such as on reaching hurts in (2)), but where the 
assistance of a higher level resolver(to use Abney's 
terminology (1987, 1989)), is required to solve 
the difficult cases, (such as on reaching melted in 
(1)). This "core parser" has been the subject of 
a number of computational implementations, in- 
cluding Marcus's deterministic parser (1980), De- 
scription theory (henceforth, D-theory) (Marcus 
et al (1983)), and Abney's licensing based model 
(1987, 1989). It has also been the subject of a 
number of psycholinguistic studies on a more the- 
oretical level (Pritchett (1992), Gorrell (in press)). 
The implementation described in this paper is 
based on the most recent model, that of (Gorrell 
(in press)). This model is interesting in that it 
does not allow the parser to employ delay tactics, 
such as using a lookahead buffer (Marcus (1980), 
Marcus et al (1983)), or waiting for the head of 
a phrase to appear in the input before construct- 
ing that phrase (Abney (1987, 1989), Pritehett 
(1992)). Instead, processing is guided by the prin- 
ciple of Incremental Licensing, which states that 
"the parser attempts incrementally to satisfy the 
principles of grammar". For the purposes of this 
implementation, I have interpreted this to mean 
that each word must be attached into a fully- 
connected phrase marker as it is found in the 
input. 1 The psychological desirability of such a 
1In fact, GorreU conjectures that, where there is 
insufficient grammatical information to postulate a 
structural relation between two constituents, such as 
in a sequence of two non-case marked NPs in an En- 
glish centre-embedded construction, the parser may 
hold these constituents unstructured in its memory 
(in press, p.212). However, for the purposes of this 
implementation, we have taken the most constrained 
position. Note that, since we do not deal with such 
291 
Full Attachment model has been argued for, es- 
pecially with regard to the processing of head- 
final languages, where evidence has been found 
of pre-head structuring (Inoue & Fodor (1991), 
Frazier (1987)). Such models have also been ex- 
plored computationally (Milward (1995), Crocker 
(1991)). 
2 D-theory and Gorrell's Model 
Gorrell employs the D-theoretic device of building 
up a set of dominance and precedence relations 2 
between nodes, where the set is intended to 
be constrained by informational monotonicity, in 
that once asserted to the set, no relation may be 
deleted or overridden. Gorrell restricts this con- 
straint to Primary structural relations (i.e. domi- 
nance and precedence), while secondary relations 
(e.g. thematic and case dependencies) are not so 
constrained. Recall (2), repeated below: 
(2) John knows the truth hurts. 
At the point where John knows the truth has been 
processed, a complete clause will have been built: 
(3) Is \[NP1 John\] \[vP Iv knows\] \[NP2 the truth\]\] 
The description will include the information that 
the verb knows precedes ~IP2, and that the VP dora- . 
inates NP2. 
{...,prec(V,NP2), dom(VP, NP2), ...} 
However, on the subsequent input of hurts, the 
structure can be reanalysed by asserting an extra 
clausal node (call it S~) dominating NP2 (which 
will then become the embedded subject), but 
which is in turn dominated by the matrix VP. This 
can be achieved by adding the following struc- 
tural relations to the tree description {prec(V,S2), 
dom(VP,S2), dom(S2,NP2)} 
(4). Is \[NP, John\] \[vP Iv knows\] \[s~ \[Ni% the 
truth\] \[vP~ hurts\]\]\]\] 
Since the description before the processing of the 
disambiguating word hurts is a subset of the final 
tree description, the monotonicity requirement is 
satisfied. Note in particular, that, because domi- 
nance is a transitive relation, and because of the 
inheritance condition on trees (a node inherits 
the precedence relations of its ancestors3), the 
two statements dom(VP, NP2) and prec (V, lips) re- 
main true after reanalysis. 4 
constructions, none of the arguments presented here 
hinge on whether or not the parser may buffer mate- 
rial in this way. 
2The original D-theory model did not compute 
precedence relations, except between terminal nodes. 
3See Partee et al (1993) for a description of the 
conditions on trees, with which all tree descriptions 
must comply. 
qt will be noticed that the reanalysis here involves 
Note also that the model will correctly fail to re- 
analyse for sentence (1) above, since the reanalysis 
will require the retraction of the domination rela- 
tion between the VP of the adverbial clause and 
the NP the ice cream. 
3 Implementation 
Although Gorrell proposes a general principle to 
guide initial attachment decisions (Simplicity: No 
vacuous structure building), and specifies the con- 
ditions under which "unconscious reanalysis" may 
occur, the model leaves unspecified the problem 
of how the-system may be implemented. Of par- 
ticular interest is the problem of how the parser 
decides which relations to add to the set at each 
point in time, especially at disambiguating points. 
3.1 Lexical Representation 
The basic framework on which the implementa- 
tion is built is similar to Tree Adjoining Gram- 
mar (Joshi et al 1975). Each lexical category is 
associated with a set of structural relations, which 
determine its lexical subtree. We call this set the 
subtree projection of that lexical category. For ex- 
ample, the subtree projection for verbs in the En- 
glish grammar is as follows, where Lex is a variable 
which will be instantiated to the actual verb found 
in the input. 
{dom(S,NP), dom(S,VP), dom(VP,V), 
dom(V,Lex), prec(NP,VP) } 
Lexical categories are also associated with lists of 
left and right attachment sites. In the above case, 
NP, (which will correspond to the subject of the 
verb), will be unified with the left attachment site. 
If a transitive verb is found in the input, then the 
parser consults the verb's argument structure and 
creates a new right attachment site for an NP, 
asserting also that this new NP is dominated by 
VP and preceded by V. 
3.2 Attachment 
Simple attachment can be performed in two ways, 
which are defined below, where the term current 
tree description is intended to denote the the set 
of structural relations built up to that point in 
processing: 
Intuitively, left attachment may be thought of in 
terms of attaching the current tree description to 
the left corner of the projection of the new word, 
while right attachment corresponds to attaching 
the projection of the new word to the right corner 
a realignment of thematic and, on GB assumptions, 
case dependencies. These are examples of what Gor- 
rell calls secondary relations, which are not subject to 
the monotonicity requirement. 
292 
of the current tree description. They are equiva- 
lent to Abney's Attach-L and Attach respectively. 
DEFINITION Left Attachment: 
Let D be the current tree description, with root 
node R. Let S be the subtree projection of the 
new word, whose left-most attachment site, A is 
of identical syntactic category as R. The updated 
tree description is S LJ D, where A is unified with 
R. 
DEFINITION Right Attachment: 
Let D be the current tree description, with the 
first right attachment site A. Let S be the sub- 
tree projection of the new word, whose root R is 
of identical syntactic category as A. The updated 
tree description is S tA D, where A is unified with 
R. 
3.3 Tree Lowering 
It should be clear that, while simple left and right 
attachment will suffice for attaching arguments 
without reanalysis, it will not allow us to de- 
rive the reanalysis required in example (2). For 
this, we intuitively require some means of insert- 
ing one tree description inside another. Schemat~ 
ically, what we require is illustrated below, where 
\[1\] is intended to represent the current tree de- 
scription built up after John knows the truth has 
been parsed, and \[2\] is intended to represent the 
subtree description of the new word hurts. 
\[i\] \[2\] 
S /\ 
NP VP /\ 
V NP 
/k 
D N 
S + / \ ==> 
NP VP 
r33 
s /\ 
NP VP /\ 
V S /\ 
NP VP /\ 
D N 
We will call this operation "tree-lowering". In- 
tuitively, the operation finds a node on the cur- 
rent trec description which matches the left at- 
tachment site of the projection of the new word, 
and attaches it, while inserting the root of the new 
projection in its place. The result is that the node 
chosen is "lowered" or "subordinated". 
In order to maintain structural coherence, the new 
word attached via tree-lowering must be preceded 
by all other words previously attached into the 
description. We can guarantee this by requiring 
the lowered node to dominate the last word to be 
attached. We also need to ensure that, to avoid 
crossing branches, the lowered node does not dom- 
inate any unsaturated attachment sites (or "dan- 
gling nodes") We therefore define accessibility for 
tree-lowering as follows: 
DEFINITION Accessibility: 
Let N be a node in the current tree description. 
Let W be the last word to be attached into the 
tree. 
N is accessible iff N dominates W, and N does not 
dominate any unsaturated attachment sites. 
DEFINITION Tree-lowering: 
Let D be the current tree description. Let S be 
the subtree projection of the new word. The left 
attachment site A of Smust match a node N acces- 
sible in D. The root node R of S must be licensed 
by the grammar in the position occupied by N. 
Let L be the set of local relations in which N par- 
ticipates. Let M be the result of substituting all 
instances of N in L with R. The attachment node 
A is unified with AT. 
The updated tree-description is D U S U M s 
It will be noticed that tree-lowering is similar in 
spirit to the adjunction operation of Tree Adjoin- 
ing Grammars (Joshi et al, 1975). The difference 
is that the foot and root nodes of an auxiliary tree 
in TAG, (corresponding to the "lowered" node and 
the node that replaces it respectively) must be of 
the same syntactic category, whereas, as we have 
seen in this example, in the model proposed here, 
the two nodes may be of different categories, so 
long as the resulting structure is licensed by the 
grammar. 
In the case of example (2), at the point where the 
truth has been processed, the parser must find an 
accessible node which matches the category of the 
left attachment site of hurts (i.e. an NP). The 
only choice is NP~: 
(3) \[s \[NP1 John\] \[vv \[v knows\] \[NP~ the truth\]\] 
NOw, all the local relations in which NP2 partici- 
pates are found: 
{dora(VP,NP2), prec(V,NP~) } 
and NP2 is substituted with the root of the new 
projection, $2 to derive two new relations: 
{dom(VP, $2), prec(V, $2)} 
These relations are found to be licensed, because 
the verb which V dominates ("knows") may sub- 
categorise for a clause, so these new relations are 
added to the set 6. Now, adding the subtree pro- 
jection of hurts to the set, and unifying its left 
SNote that Abney's STEAL operation (1987, 1989) 
is more powerful than tree-lowering, since it may 
change domination relations, and thus will allow sen- 
tences such as (1), though it excludes reduced rela- 
tive garden paths, such as The horse raced past the 
barn fell. The original D-theory model (Marcus et al 
(1983)) is also more powerful, because it allows the 
right-most daughter of a node to be lowered under a 
sibling node. 
eNote that the relations defining the original posi- 
tion of NP2, (i.e. dom(l/P,NP2) and prec(V,NP2)) are 
not subtracted from the set. 
293 
attachment site with NP2 results in the derived 
structure with NP~ "subordinated" into the lower 
clause. 
Is \[NP~ John\] \[vp Iv knows\] Is2 \[gP2 the truth\] 
\[vP~ hurts\]\]\]\] 
With the tree-lowering operation so defined, the 
problem of finding which relations to add to the 
set at a disambiguating point reduces to a search 
for an accessible node at which to apply this oper- 
ation. However, this implies that, if more than one 
such node exists, the parser must be given a pref- 
erence for making the requisite decision. Consider 
the following sentence fragment, for example: 
(5) I know \[NP1 the man who believes \[NP2 the 
countess\]\]... 
If the input subsequently continues with a verb, 
then we have a choice of two nodes for lower- 
ing, i.e. NP1 and l~P2. Though no experimental 
work has been done on this type of sentence, there 
seems to be an intuitive preference for the lower 
attachment site, NP2. In (6), binding constraints 
force lowering to be applied at NP2, while in (7), it 
must be applied at NP1. Of the two, most native 
English speakers report (6) to be easier. 
(6) I know the man who believes the countess 
killed herself. 
(7) I know the man who believes the countess 
killed himself. 
Note also, that, on standard X-bar assumptions, 
the attachment of post-modifiers may be derived 
via lowering at an X I node. In this case, the low- 
ered node and its replacement will be of the same 
syntactic category (like the root and foot node of 
a TAG auxiliary tree). Researchers have noted 
a general preference for low attachment of post- 
modifiers (this is accounted for by the principle 
of late closure (Frazier and Rayner, 1982)). This 
would suggest that a reasonable search strategy 
for English would be to search the set of accessi- 
ble node in a bottom-up direction for English. 
The algorithm is constructed in such a way that 
lowering is only attempted in cases where sim- 
ple attachment fails. This means that arguments 
(which are incorporated via simple attachment) 
will be attached preferentially to adjuncts (which 
are incorporated via lowering). This captures the 
general preference for argument over adjunct at- 
tachment, which is accounted for by the princi- 
ple of Minimal attachment in Frazier and Rayner 
(1982), and by the principle of simplicity in Got- 
tell (in press). 
4 Processing Japanese 
4.1 Main/subordlnate clause ambiguity 
Japanese presents a challenge for any incremental 
parsing model because, typically, it is not possible 
to determine where an embedded clause begins. 
Consider the following example: 
(8) John ga \[Oi ronbun wo kaita\] seitoi wo hometa. 
John NOM essay ACC wrote student ACC praised 
"John praised the student who wrote the essay" 
Up to the first verb kaiia ("wrote"), the string 
is interpretable as a full clause (without a gap), 
meaning '~John wrote an essay", and the incre- 
mental parser builds the requisite structure. How- 
ever, the appearance of the head noun seito (stu- 
dent) means that at least part of the preceding 
clause must be reinterpreted as a relative clause 
including a gap (note that there is no overt relative 
pronoun in Japanese). One way of looking at what 
is happening here is to see the subject NP John 
ga as being dissociated from the clause in which 
it is originally attached, and reattached into the 
main clause. But looking at it from a different 
perspective, as Gorrell has noted (in press), one 
can see the subject NP as remaining in the main 
clause, and the constituent bracketed in (8), (ton- 
bun wo kaita ("wrote an essay")) as being lowered 
into the relative clause. If this is possible, then we 
would expect examples like (8) to be unconscious 
garden paths, and this does indeed seem to be re- 
flected in the intuitive data (see Mazuka and Itoh 
(in press)). However, if we are to allow our parser 
to handle such examples, we must expand the def- 
inition of tree-lowering, since, in order to build a 
relative clause, we have to assert extra material 
(including the empty subject and the new S node), 
which is not justified solely by the lexical require- 
ments of the disambiguating word, the head noun 
seito. This involves reconstructing all the clausal 
structure dominating the lowering site (including 
asserting empty argument positions), with refer- 
ence to the verb's case frame, and attempting to 
attach the result as a relative clause to the head 
noun. 
4.2 Minimal Expulsion 
Inoue (1991), describes a "minimal expulsion 
strategy", which predicts a preference, on reanal- 
ysis, towards expelling the minimum amount of 
material from the clause. In our terms, this means 
that (assuming a binary right-branching clause 
structure, with the verb in its right corner) the 
node selected for lowering must be as high as 
possible. This means that the bottom-up search 
which we use for English will wrongly predict 
a Maximal expulsion strategy. In cases such as 
(8), assuming the bottom-up search, when a post- 
clausal noun has been reached in the input, the 
294 
parser starts its search from the node immediately 
dominating the last word to be incorporated, (i.e. 
the verb of what will become the relative clause). 
This means that, in cases such as (8), the first 
preference will be to lower the verb (and therefore 
"expel" both subject and object), whereas the hu- 
man preference, (to lower the object and verb, and 
therefore expel only the subject) is the parser's 
second choice on the bottom-up search strategy. 
Mazuka and Itoh (in press) note that examples 
where both subject and object must be expelled 
from the relative clause, as would be the first 
choice in a bottom-up search, often cause a con- 
scious garden path effect. An example, adapted 
from Mazuka and Itoh is the following: 
(9) Yamasita ga yuuzin wo \[O Oi houmonsita\] 
kaisyai de mikaketa. 
Yamasita NOM friend ACC visited company 
LOC saw 
"Yamasita saw his friend at the company he vis- 
ited." 
In order to capture the minimal expulsion strat- 
egy in this class of Japanese examples, therefore, 
search for the lowering node should be conducted 
top-down. We are currently investigating the con- 
sequences of changing the search strategy in this 
way. 
5 The Problem of Retrospective 
Reanalysis 
Having formulated the constraints of Gorrell's 
model in terms of the accessibility of a node for 
~ree-lowering, we can see that the model can be 
falsified if we can find a case where the relevant 
disambiguating information comes at a point in 
processing where the node which is required to 
be lowered is no longer accessible. Consider the 
following pair of sentences: 
(10) I saw the man with the moustache. 
(11) I saw the man with the telescope. 
It is familiar from the psycholinguistic literature 
that there is a preference for attaching the with 
phrase as an instrumental argument of the verb 
(as in (11), on the reading where the telescope 
is the instrument of seeing). On the assumption 
that saw selects for a PP instrumental argument, 
we can derive this preference in the present model 
via the preference to attach as an argument as op- 
posed to an adjunct. However, since we are con- 
strained by incrementality, we will have to make 
an attachment decision for the PP as soon as the 
preposition with is encountered, and it will be at- 
tached in the preferred reading as a sister of the 
verb. This means that, in cases such as (10), 
where, on the globally acceptable reading, the PP 
is an adjunct of the NP the man, this attachment 
will have to be revised, and the PP retrospectively 
adjoined into the relevant N t node. However, once 
the preposition with has been attached, the re- 
quired N' node will no longer be accessible, and 
a conscious garden path effect will be predicted, 
which, intuitively, does not occur. Note that there 
is no garden path effect even if the preposition is 
separated from the disambiguating head noun by a 
series of adjectives: ("I saw the man with the neat, 
quaint, old-fashioned moustache/telescope"). 
The same result obtains if we abstract away from 
the particular implementational details of tree- 
lowering, and return to the abstract level at which 
Gorrell states his model. Once the PP has been 
attached as an argument of the verb, it can never 
be reanalysed as the adjunct of the preceding NP, 
because the NP will precede the PP before re- 
analysis, and dominate it after reanalysis, which 
is against the "exclusivity condition" on trees (i.e. 
no two nodes may stand in both a dominance and 
a precedence relation). 7 
A similar problem concerns examples such as the 
following, from Gibson et al (1993): 
(12) the lamps near the 
\[that was damaged in the 
(13) the lamps near the 
paintings of the house 
flood\]. 
painting of the houses 
\[that was damaged in the flood\]. 
(14) the lamp near the paintings of the houses 
\[that was damaged in the flood\]. 
in the above, Gibson et al have manipulated num- 
ber agreement to force low (12), middle (13) and 
high (14) attachment of the bracketed relative 
clause. The results of their on- and off-line ex- 
periments show clearly that the low attachment 
(corresponding to 12) is easiest, but the middle 
attachment (corresponding to (13)) is most diffi- 
cult. This behaviour cannot be captured whether 
we adopt a bottom-up or a top-down search for 
tree-lowering. However, even if we can incorpo- 
rate the required preferences into the parser, the 
constraint of incrementality will force us to make 
the decision on encountering that. This means 
that, assuming we decide initially to attach low, 
but number agreement on was subsequently forces 
high attachment, as in (14), then a conscious gar- 
den path effect will be predicted, as lowering can- 
not derive the reanalysis. This is true on the ab- 
stract level as well, since there will be nodes in 
the description which precede the original low po- 
sition of the relative clause, but are dominated by 
the subsequent high position of the relative clause. 
ZNote that in Marcus et al (1983), since precedence 
relations were not computed for non-terminals, lower- 
ing into a predecessor was possible, thus (11) would 
cause no processing difficulty. However, presumably, 
their parser would overgenerate on examples such as 
the horse raced past the barn fell. 
295 
However, intuitively, of the above sentences, it is 
only (13) which causes the conscious garden path 
effect.S 
6 Conclusion 
The current implementation shows that the suc- 
cess of an abstract model such as Gorrell's de- 
pends crucially on the computational details of 
the processing algorithm used. The search for the 
lowering site is of particular importance. In the 
final section we have seen that the combination of 
informational monotonicity with the assumption 
of strict incrementality results in a system which 
is too constrained to capture all the processing 
data. Future research will be aimed at determin- 
ing, firstly, how we can enrich the information to 
which the search strategy is sensitive in order to 
provide a better match with human preferences, 
and secondly, which constraints should be relaxed 
in order to avoid the problem of undergeneration. 
References 
Abney, S. P. (1987): Licensing'and Parsing. Pro- 
ceedings of NELS 17 p.l-15, University of Mas- 
sachusetts, Amherst 
Abney, S. P. (1989): A computational model of 
human parsing. Journal of Psycholinguistic Re- 
search 18 p.129-144 
Crocker, M. W. (1991): A Logical Model of Com- 
petence and Performance in the Human Sentence 
Processor. PhD thesis, Dept. of Artificial Intelli- 
gence, University of Edinburgh, Edinburgh, U.K. 
Frazier, L. (1987): Syntactic processing: Evidence 
from Dutch. In Natural Language and Linguistic 
Theory 5.4 p.519-559 
Frazier, L. and K. Rayner, (1982): Making and 
correcting errors during sentence comprehension: 
Eye movements in the anMysis of structurally 
ambiguous sentences. Cognitive Psychology 14 
p.178-210 
Spreliminary findings suggest that a similar pref- 
erence rating is employed in (written) production as 
well as (reading) comprehension for these examples. 
This can be seen in Gibson et al's (1994) study. This 
shows a the LOW> HIGH > MID ordering in the at- 
tachment of the final PP in NPs of the following form 
found in the Brown corpus: 
NP1 Prep NP2 Prep NP3 PP 
Of 105 unambiguous PP adjunct attachments, 68% 
were low-attached, 26% high attached and 10% mid- 
attached. However, the question of whether the syn- 
tactic structures people preferentially use in produc- 
tion should correspond to the syntactic structures 
people preferentially assign to strings during compre- 
hension is still very much an open issue, though see 
Mitchell and Cuetos (1991) for a view that the expe- 
rience of previous input influences parsing decisions. 
Gibson, E., N. Pearlmutter, E. Canesco-Gonzalez 
and Greg Hickok (1993): Cross-linguistic Attach- 
ment Preferences: Evidence from English and 
Spanish. (ms. submitted to Cognition) 
Gibson, E. and N. Pearlmutter, E. (1994): A 
corpus-based account of Psycholinguistic Con- 
straints on Prepositional Phrase Attachment (in 
C. Clifton, L. Frazier and K. Rayner (eds) 
Perspectives on Sentence Processing New York: 
Lawrence Erlbaum 
Gorrell, P. (in press): Syntax and Perception. to 
be published by Cambridge University Press 
Inoue, A. (1991): A comparative study of parsing 
in English and Japanese. PhD thesis, University 
of Conneticut. 
Inoue, A. and J.D. Fodor (in press): Information- 
paced parsing of Japanese. (to appear in Mazuka 
~; Nagai (eds)) 
Joshi, A.K., L.S. Levy, and M. Takahashi, (1975): 
Tree Adjunct grammars. Journal of Computer 
and System Sciences 10, p.136-163 
Marcus, M. (1980): A Theory of Syntactic Recog- 
nition for Natural Language Cambridge, MA: 
MIT Press 
Marcus, M., D. Hindle, and M. Fleck (1983): D- 
theory: Talking about talking about trees. As- 
sociation for Computational Linguistics 21 p.129- 
136 
Mazuka, R. and K. Itoh (in press): Can Japanese 
be led down the garden path? (to appear in 
Mazuka and Nagai) 
Mazuka, R., and Nagai (eds)(to appear): 
Japanese Syntactic Processing Hillsdale, NJ: 
Lawrence Earlbaum 
Milward, D. (1995): Incremental Interpretation 
of Categorial Grammar. in Proceedings of EA CL 
(this volume) 
Mitchell, D.C. & Cuetos, F. (1991): The origins of 
parsing strategies. Conference proceedings: Cur- 
rent issues in natural language processing Univer- 
sity of Texas at Austin, TX 
Partee, B., A. ter Meulen and R. E. Wall (1993): 
Mathematical methods in Linguistics Dordrecht: 
Kluwer Academic Publishers 
Pritchett, B. L. (1992): Grammatical Competence 
and Parsing Performance. Chicago, IL: Univer- 
sity of Chicago Press 
296. 
