|Date: Mon, 25 Aug 2003 08:38:47 +0200
From: Michael Moss
Subject: Logical Form and Language
Preyer, Gerhard and Georg Peter, ed. (2002) Logical Form and Language,
Oxford University Press
Michael Moss, Ph.D., University of Gdansk, Department of General and
This book is a collection of 17 essays by 21 authors. Each article deals
with a different aspect of what Logical Form is and how it functions
together with language. Topics range from how Russellian Logical Form can be
made to deal with quantification more elegantly (Neale, pp.13-53), to a
detailed analysis of thematic roles as absolutes in verbal arguments, to a
compromise between Russellian and Millian analyses of proper names. Although
this is not an introduction to Logical Form, each of the articles gives
enough background to their individual problems that one can follow the
arguments without being a specialist. At the same time, the articles are at
a technical level which will satisfy those who are more initiated.
Furthermore, the range of topics covered in the book gives a good overview
of what kind of work is being done in Logical Form today.
After a short introduction by the editors, the book is divided into three
section: 1. The Nature of Logical Form, 2. Intentionality Events and
Semantic Content, and 3. Logical Form, Belief Ascription and Proper Names.
Each of the articles in its own way places the study of language and logic
into the larger picture of philosophy of language and the problem of how
natural language helps us to interpret and communicate information about the
real world. Interestingly, two topics continued to surface throughout the
book, namely "How are the ideas communicated through language 'bound' to
items in the real world?"; and "How do Logical Form, Semantics and Syntax
interact and where are the borders between them?" The first of these
questions is interesting due to the fact that Binding and co-indexing are
currently out of vogue in generativist circles, and the second because of
the Minimalist Program's desire to somehow reconcile the interface between
LF and Semantics (see particularly Chomsky 2001, 2002).
This volume of essays tries to fill the role of a handbook of current work
in Logical Form. The book is quite broad in scope, which is good for the
general reader who wants to know more about current research in Logical Form
and the problems being addressed there, but perhaps less interesting for one
who is currently involved in such research. On the other hand, each of the
articles is quite technical, making the book much more than an 'introduction
As mentioned above, the book contains 17 individual articles, including the
introduction. In the interest of length and considering my personal areas of
competence, I have chosen 4 articles to comment on in more detail. This
selection does not raise these articles above the rest in any way, except
that they were closer to my area of research and thus it is easier for me to
comment on them. I will discuss the articles by: Fiengo and May,
Higginbotham, Hornstein, and Schein.
"Identity Statements" by Fiengo and May (pp. 169-203) investigates the
Logical Form of identity statements such as: "Tully is Tully" vs. "Tully is
Cicero". The first is generally understood as a truism, but the second is
not, although intuitively the statements are understood to be related. Frege
presented two analyses of this problem, the first says that two identity
statements can have the same reference, while the second says that identity
statements present the 'same' information in two different ways
(technically, that the verb 'to be' is understood to refer to objectual
identity). Fiengo and May point out that both of Frege's analyses are
insightful and that their solution to the problem will allow both to be
correct. As they show, the truism 'A is A' is 'meaningless' in the obvious
situation. However, their example "But Max, Paderewski is Paderewski", where
the speaker is explaining to Max that he was mistaken in thinking that
Paderewski the pianist and Paderewski the statesman were two different
people, shows that 'A is A' type sentences can in fact be informative. They
propose that if the statesman and the pianist are perceived as two
individuals, then the statement should actually be read as "But Max,
Paderewski(1) is Paderewski(2)", which makes the sentence more like a
statement of the type: A=B and not A=A.
It seems that this type of interpretation has interesting implications for
co-indexing and binding theories. The indexes used here are not part of the
syntax, as co-indexing was proposed in the Government and Binding model, but
are part of the Semantics. This leads to questions as to whether or not
binding is part of the syntax or part of semantics and whether the indexing
used by the Semantic component is visible at the level of Syntax. Perhaps
'identity' can be seen as a feature, which would then allow it to work with
the syntax for binding purposes using an 'Agree' type mechanism.
Higginbotham's article ''Why is Sequence of Tense Obligatory'' (pp. 207- 227)
investigates the consequences of treating Tense as an anaphoric relation
holding between events. As with other binding phenomena then, Tense can
also be seen to be ambiguous. For instance, a sentence such as ''John said
that Mary was pregnant'' is ambiguous about the time when Mary was actually pregnant.
Higginbotham proposes that this is due to the fact that Tense can be interpreted
two (or more) ways similarly to ambiguous binding relations depending on
whether it is +past or -past. That is, in Higginbotham's analysis, the event in
the subordinate clause can be both anaphoric and linked to the
speaker's utterance giving the so-called "double-access" sentences.
I shall not criticize this argument, but would like to comment on its value. By using
Davidson's event structure in Logical Form, and the familiar understanding of
anaphora from syntactic theory, Higginbotham is able to put forth some interesting
ideas about how Tense 'works' without falling back on metaphysical arguments.
This seems to be a significant step in the right direction.
Next I comment on Barry Schein's extensive article "Events and the Semantic
Content of Thematic Relations". Here Schein attempts to resolve the
problem for Davidsonian events in 'plural' environments such as Carnegie
Hall sitting opposite the Carnegie Deli and the Carnegie Deli sitting
opposite Carnegie Hall. While the sentences give both NPs different thematic
roles depending on the grammatical function in the sentence, the logical
intuition is that both objects can be seen to fill the same relative role in
the real world. Schein proposes that Davidson's term 'event' may be
confusing and that a better term would be 'scene' so that two individual
events could be derived from one scene. The important question here is
whether thematic roles can be seen as absolute in terms of syntactic
structure and Logical Form. Schein proposes that, in fact, they can, if the
syntax has appropriate zero morphemes located in a higher functional
projection. This is a very interesting proposal, and the article contains a
detailed discussion of prepositional phrases and their relation to thematic
roles in Logical Form.
Finally, I consider Hornstein's article "A Grammatical Argument for a
Neo-Davidsonian Semantics". Here, Hornstein shows that the elimination of
chains from syntactic structures "requires" a Davidsonian type of logical
form, which makes explicit how the thematic roles are distributed by
illustrating the relation of the various arguments to the event. Not only
does this argument reduce the amount of syntactic machinery (by eliminating
chains and replacing it with a copy based theory), but it also gives
evidence for thematic roles being feature based, which gives them a much
firmer setting in syntactic theory.
I have commented on the above articles because they were interesting to me
personally. All of the remaining articles present their arguments with at
least as much detail and care. The scope of the book is far-reaching, but
each of the articles has enough room to give the uninitiated reader the
necessary background to understand the content.
Chomsky, Noam 2001. "Derivation by phase". In Michael Kenstowicz (ed.) 2001.
Ken Hale: a life in language. Cambridge Mass: MIT Press. pp. 1-52.
Chomsky, Noam 2002. "Beyond explanatory adequacy". MIT Ocassional Papers in
Linguistics 19. Cambridge, Mass.:MITWPL.