Substitutional Quantification and Quotation

Author(s): Gilbert Harman

Source: Noûs, Vol. 5, No. 2 (May, 1971), pp. 213-214
Published by: Wiley
Stable URL: http://www.jstor.org/stable/2214733 .
Accessed: 14/06/2014 15:49

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Wiley is collaborating with JSTOR to digitize, preserve and extend access to Noûs.

http://www.jstor.org

This content downloaded from 195.34.79.223 on Sat, 14 Jun 2014 15:49:48 PM

All use subject to JSTOR Terms and Conditions
 Quanticationand Quotation
Substitutional

GILBERT HARMAN
PRINCETON UNIVERSITY

Robert Binldey points out' that substitutionalquantification

(as opposed to referentialor objectualquantification)ought to per-
mit the formulationof such statementsas
(1) ( 3 n) ("n"has n letters)
"where 'n' is understood to range substitutionally over English
numberwords,"so that (1) is "shownto be true"by e.g.
(2) "Four"has four letters.
He notes the following difficulty with this line of thought.
If there are no restrictions on substitutional quantification and
quotation, semantical notions like truth and denotation become
definable in the object language so that the semantic paradoxes
result.
(3) x is true = df ( 3 s) (x = "s" & s).
(4) x denotes y= df (3 n)(x="n"&n=y).
He thereforeintroducesthe following "no-mixingrestriction":
"a single quantifier cannot simultaneously bind variables both
within and without the quotation."The right hand sides of (3) and
(4) violate the no-mixing restriction,so truth and denotation are
not definable as indicated. Unfortunately,(1) also violates the no-
mixing restriction.
A natural alternative,which preserves (1), is to introduce a
cumulative hierarchy of metalanguages and to suppose that the
1 Robert Binkley. "Quantifying, Quotation, and a Paradox," Nods IV
(1970): 271-277. My research is supported by NSF.
213

This content downloaded from 195.34.79.223 on Sat, 14 Jun 2014 15:49:48 PM

All use subject to JSTOR Terms and Conditions
214 NONS

quotationof an expressionin language L is not an expressionin L

but only in L + 1 and higher languages.Quotationmarkscan carry
subscriptsindicating the language to which the expressionquoted
belongs. Truth and denotationfor expressionsin L would be defin-
able:
(5) x is true = df ( 3 s) (x = L S"L & S).
(6) x denotes y = df ( 3 n)(x = "Ln"L&n = y).
But they are not definablein language L itself, since the right hand
sides of (5) and (6) are not sentences of L but only of L + 1 and
higher languages.
Similarly (1) remains. More precisely,
(7) (3 n) ("Ln"L has n letters)
is well formedfor any choice of L.

This content downloaded from 195.34.79.223 on Sat, 14 Jun 2014 15:49:48 PM

All use subject to JSTOR Terms and Conditions