357

HOW TO MAKE THE DISTINCTION BETWEEN CONSTATIVE

AND PERFORMATIVE UTTERANCES
B Y KLAUS H. JACOBSEN

In his paper " Performatif-Constatif " presented at Royaumont in 19581

J. L. Austin argues that there is no clearcut distinction between performative

Downloaded from http://pq.oxfordjournals.org/ at KTH Royal Institute of Technology on March 9, 2015

and constative utterances. A great deal of the literature concerning per-
formatives has, since the publication of Austin's William James Lectures
How to Do Things with Words (Cambridge, Mass., 1962) and the above-
mentioned paper, paid lip-service to this withering away of the dichotomy
between performative and constative utterances. 2 In opposition to this
viewpoint, I should like to argue that a very simple and straightforward
device will suffice to show that the dichotomy can be made distinct, and
indeed has to be, if one does not wish to join Austin in his relativism which,
as I shall show, is the consequence of the reduction of all statements to
performatives.
As far as Austin is concerned, a performative utterance is neither true
nor false, i.e. it is neither true nor false to me or anybody else that I say,
for instance : " I apologize for my behavior ". It may be right or wrong
for me to do so, because the utterance ' I apologize ' is used to perform
rather than to describe the act. Being neither true nor false the performative
utterance may, however, be " abused " in three different ways, which
Austin now and then terms " unhappiness " of the performative formula.
First, the performative may be " null and void ", if it is issued by a person
who is in no position whatever to perform the act in question, i.e. who is
not authorized either specifically by a group of persons, or in general by
some legal right. Second, the performative may be issued with no degree
of intention whatever to fulfil the appropriate requirements of the act
performed. Third, the performative utterance may be issued with varying
degrees of intention (breach of commitment). ' I welcome you, but I am
going to bed ' may be an example of the third type of unhappiness. Having
circumscribed performatives in this way Austin proceeds to show that even
though performative utterances generally speaking have the grammatical
form of an utterance in the first person singular followed by a verb in the
present indicative active, or in the second and third person followed by a
verb in present indicative passive, as e.g. ' You are requested to shut the
door ', no specific grammatical form will suffice to distinguish between per-
formative utterances and other possible types of utterance such as constative
1
" Performatif-Constatif ", in La Philosophic Analytique (Paris, 1962), pp. 271-304.
Translated as " Performative-Constative ", in Charles E . Caton (ed.), Philosophy and
Ordinary Language (Urbana, Illinois, 1963), pp. 22-33. References are to this translation.
a
A good example is J . Hartnack, " Performative Utterances ", in P. Edwards (ed.),
The Encyclopaedia of Philosophy (London and New York, 1967).
358 KLAUS H . JAOOBSBN

ones—or, as Austin says : " We can't, then, expect any purely verbal cri-
terion of the performative ". 3
Without a purely verbal criterion of performative utterances it becomes
a question of importance whether performatives and performatives alone
are liable to the unhappinesses which Austin has listed. If it can be demon-
strated that constative utterances likewise are liable to a familiar type of
unhappiness, the dichotomy between constative and performative utterances
seems unclear.

Downloaded from http://pq.oxfordjournals.org/ at KTH Royal Institute of Technology on March 9, 2015

As is well known, Austin does show rather successfully4 that normally
agreed upon constative statements such as ' All John's children are bald,
but (or ' and ') John has no children ' and ' The cat is on the mat, but (or
' and') I don't believe it is ', and ' All the guests are French, and some of
them aren't' are statements equally as unhappy with respect to the notions
of presupposition, implication and entailment (respectively) as performative
utterances may be with respect to nullity, insincerity and breach of commit-
ment (degree of intention). This analysis brings us to the tempting conclusion
that since there is no purely verbal criterion by which to distinguish perfor-
mative utterances from other types of utterance, and since constative and
performative utterances share a similar fate of infelicity, constative state-
ments are nothing but a breed of performative utterance. This, however,
to me is simply untrue.
The argument advanced by Austin presupposes either an exhaustive
classification of all qualities belonging to performative and constative utter-
ances or at least a classification of the essential qualities of both classes.
Austin admits that his investigation falls short of the first requirement de
facto, but I am sure he is convinced in his paper that he has shown what are
at least the essential qualities characterizing both performative and con-
stative utterances. I think he fails to show this, and will suggest one essential
quality which Austin overlooked when he wrote his paper, and which to
the best of my knowledge has been ignored by anyone writing on the subject
since.
This quality, which can be ascribed to constative utterances but not
to performative ones, draws its strength from the notion of inter-personal
transference of a constative utterance. Thus statements such as ' The cat
is on the m a t ' are statements which can be transferred from one person to
another without any prior agreement having been made. I am for instance
as free to use the statement ' The cat is on the m a t ' as you are, but it is
not the case that I am free to use, i.e. to take over, an act which you by
means of a performative utterance have performed. If you perform an act
of promising by way of the performative ' I promise you ', this act cannot
be taken over by me without us making an agreement whereby you give
me the authority to act on your behalf and whereby I, by means of at least
H)p. cit., p . 26.
K>p. cit., 27-8.
CONSTATIVE AND PERFORMATIVE UTTERANCES 359

one performative utterance, have accepted your obligation(s). The intrinsic

quality of a performative utterance seems indeed inseparably connected
with the fact that an act is either performed explicitly by a person, as e.g.
' / promise you ', or by a group of persons authorized to act as one person,
as e.g. ' You are requested to shut the door ', where the request has to be
issued either by one person or by a group of persons (officials), who have
the authority to make requests concerning the regulations governing the
public behavior. In short, an act is always an act performed by somebody.

Downloaded from http://pq.oxfordjournals.org/ at KTH Royal Institute of Technology on March 9, 2015

In this respect it is of course true that any speech act is an act performed
by somebody, but from this obvious fact it does not follow that I perform
an act by what I say. It only follows that in saying something I perform
the act of saying. From the fact that an act belongs to somebody, to a
person, it becomes clear that the performative utterance being a specific
act does belong specifically to the person uttering it, whereas the constative
statements do not belong especially to any person, even though both types
of utterance viewed as speech acts are acts. Hence without any preceding
procedure the performative utterances are non-transferable and the con-
statives are transferable.
Now, the question is whether this notion of transference enables us to
make a sharp distinction between performative and constative utterances.
Are there any utterances which may be classified as non-transferable con-
statives and are there any transferable performatives ? The first may seem
to be the case. A statement such as ' I am a Dane ' purports to describe
something, i.e. it is a constative statement, but it does not seem transferable
in quite the same way as e.g. ' The cat is on the m a t ' . This, however, applies
to all constative statements expressing the logical subject by means of a
first person singular personal pronoun. The point to be made in order to
cope with this class of utterances is relatively simple but extremely im-
portant. An utterance like ' I am a Dane ' is transferable because any name
of a person to whom the ' I ' refers could be substituted for the ' I ' , resulting
in a transferable statement. For example, if the ' I ' refers to Mr. Jensen,
then ' Mr. Jensen is a Dane ' is a statement equivalent to the statement
' I am a Dane ' as uttered by Mr. Jensen, but which could obviously be
uttered equally by anyone else. Thus ' I am a Dane ' is a statement which
may be stated by Mr. Jensen and Mr. Andersen without being specifically
tied to any one of these persons, while a statement such as ' I promise you
my tape-recorder ' is tied up with the person making the promise to an
identifiable other person, and hence is not transferable without a specific
procedure. In the case of ' I promise you my tape-recorder ' the substitution
of the name of the person to whom the ' I ' refers gives either a non-transfer-
able performative, or a non-equivalent constative. ' Mr. Jensen promises
you his tape-recorder ' might express a promise when uttered by Mr. Jensen,
but can only describe a promise when uttered by anyone else (unless they
have been authorized to act on Mr. Jensen's behalf).
360 KLATJS H . JACOBSETf

Having dealt at least preliminarily8 with apparently problematic con-

stative utterances it remains to be shown that explicit performative utter-
ances are not transferable without a preceding procedure whereby the act
and the responsibility of the act undertaken by the agent is transferred to
another person. Take for instance the utterance : ' I warn you that the
bull is dangerous '. Is this a non-transferable utterance or is it the case
that it can be transferred as easily as the problematic constative state-
ments ? My answer runs in favour of the first alternative. In the utterance

Downloaded from http://pq.oxfordjournals.org/ at KTH Royal Institute of Technology on March 9, 2015

' I warn you that so-and-so ' the ' I ' in question may potentially represent
anybody placed in appropriate circumstances to issue warnings, but in the
very moment it is uttered by a person it has become an act performed by
a specific person. This act is tied to this person in such a way that the
responsibility for being right or wrong in warning somebody is placed
wholly upon the person performing the act of warning. His and his alone
is the responsibility, and his and his alone was the act producing this re-
sponsibility. Therefore only a procedure of transference can bestow validity
upon a substitution of the agent. The dichotomy between constative and
performative utterances based upon the notion of transference is accordingly
defendable and clear-cut, provided that the connection between the ' I '
and the act performed verbally by the speaker is intrinsically different from
the connection between the ' I ' and the description verbally made by him.
If one is not willing to accept this, the dichotomy breaks down, and I
strongly believe as presumably did Austin that it breaks down in favour of
a reduction of all constatives to performatives. However, the consequence
of this should by now be clear. If one does not accept the point that the
role played by the ' I ' is fundamentally different with respect to the descrip-
tion of an act from what it is with respect to the performance of an act, this
leads us to the conclusion that a descriptive statement such as ' The cat is
on the m a t ' must have a form like : ' I who am placed in appropriate circum-
stances state the following : The cat is on the m a t ' . This utterance being
a performative is neither true nor false but it may of course be unhappy in
different ways. Depending on the person uttering it and the totality of his
situation it may also be right or wrong. In short, the " truth " of the utter-
ance is relative to the person uttering it and relative in such a way that
" man is the measure of all things, of those that are that they are, of those
that are not that they are not ". The obliteration of the distinction between
constative and performative utterances apparently brings us in the company
of Protagoras. This of course is not bad company, but if one fails to notice
this consequence, as I think Austin did, one equally fails to enjoy the
dubious pleasures of relativism.
University of Odense.
"The word " preliminarily " indicates in this connection t h a t I do not exclude the
possibility of other existing types of problematic constative utterance. I t may be argued
for instance t h a t egocentric psychological descriptions constitute such a type. I n my
opinion, however, the problems concerning the transferability of these can be solved
along the same lines as I have proposed for the problematic statements treated in this
paper.