Intro and study questions for Goffman's "Footing"

You only need to read sections i, vi, and viii through xi (1, 6, 8-11).

Goffman is probably the most influential historical figure in the study of social
interaction in America. Not only was his research influential, but he was an influential
teacher, shaping the research of graduate students who went on to become influential
scholars and teachers themselves. In this article, Goffman addresses the ways in which
interlocutors build constantly shifting, variously embedded activities in their talk and
interaction. Shifting with these shifting activities are the projected perspectives and roles
of various participants in the social interaction.

We are reading this because it gives you perspectives and analytical categories that you
can use in your analyses of social interaction for your upcoming assignments. In your
next assignment, for example, you will be asked to isolate the boundaries of an activity in
talk and interaction. The boundaries of discrete activities in talk often coincide with
changes in footing.

1. What is "footing"?

2. How do monolingual speakers commonly signal that a change in footing is occurring?

(p. 128)

3. What is a "coordinated task activity", p. 143, and what is distinctive about it, according
to Goffman, when compared to other forms of conversation?

4. What is "embedding"? (section viii)

5. What frame, or footing, was embedded in the news conference by Helen Thomas'
pirouette? (p. 156)
FOOTING

Consider a journalistically reported strip of interaction, a news

bureau release of 1973 o n presidential d0ings.l The scene is the
O v a l Office, the participants a n assemblage of government offic-
ers a n d newspaper reporters gathered in their professional capaci-
ties for a political ritual, the witnessing of t h e signing of a bill:
WASHINGTON [UPII-President Nixon, a gentleman of the old
school, teased a newspaper woman yesterday about wearing slacks
to the White House and made it clear that he prefers dresses on
women.
After a bill-signing ceremony in the Oval Office, the President
stood up from his desk and in a teasing voice said to UPI's Helen
Thomas: "Helen, are you still wearing slacks? Do you prefer them
actually? Every time I see girls in slacks it reminds me of China."
Miss Thomas, somewhat abashed, told the President that Chi-
nese women were moving toward Western dress.
"This is not said in an uncomplimentary way, but slacks can do
something for some people and some it can't." He hastened to add,
"but I think you do very well. Turn around."
As Nixon, Attorney General Elliott L. Richardson, FBI Director
Clarence Kelley and other high-ranking law enforcement officials
smiling [sic], Miss Thomas did a pirouette for the President. She
was wearing white p a ~ t sa, navy blue jersey shirt, long white beads -
and navy blue patent leather shoes with red trim.

I . Grateful acknowledgment is made to S~miotica,where this paper first

appeared (25[1979]:1-29).
 Footing
Nixon asked Miss Thomas how her husband, Douglas Cornell,
liked her wearing pants outfits.
"He doesn't mind," she replied.
"Do they cost less than gowns?"
"No," said Miss Thomas.
"Then change," commanded the President with a wide grin as
other reporters and cameramen roared with laughter. [The Evening
Bulletin (Philadelphia), 19731

This incident points to the power of the president to force an

individual who is female from her occupational capacity into a
sexual, domestic one during an occasion in which she (and the
many women who could accord her the role of symbolic repre-
sentative) might well be very concerned that she be given her full
professional due, and that due only. And, of course, the incident
points to a moment in gender politics when a president might
unthinkingly exert such power. Behind this fact is something
much more significant: the contemporary social definition that
women must always be ready to receive comments on their "ap-
pearance," the chief constraints being that the remarks should be
favorable, delivered by someone with whom they are acquainted,
and not interpretable as sarcasm. Implied, structurally, is that
woman must ever be ready to change ground, or, rather, have
ground changed for her, by virtue of being subject to becoming
momentarily an object of approving attention, not-or not
merely-a participant in it.
The Nixon sally can also remind us of some other things.
In our society, whenever two acquainted individuals meet for
business, professional, or service dealings, a period of "small
talk" may well initiate and terminate the transaction-a mini
version of the "preplay" and "postplay" that bracket larger so-
cial affairs. This small talk will probably invoke matters felt to
bear on the "overall" relation of the participants and on what
each participant can take to be the perduring concerns of the
other (health, family, etc.). During the business proper of the
encounter, the two interactants will presumably be in a more
segmental relation, ordered by work requirements, functionally
s p e s c authority, and the like. Contrariwise, a planning session
among the military may begin and end with a formal acknowl-
edgment of rank, and in between a shift into something closer
 Forms of Talk

to equalitarian decision-making. In either case, in shifting in

and out of the business at hand, a change of tone is involved,
and an alteration in the social capacities in which the persons
present claim to be active.
Finally, it might be observed that when such change of gears
occurs among more than two persons, then a change commonly
occurs regarding who is addressed. In the Nixon scene, Ms.
Thomas is singled out as a specific recipient the moment that
"unserious" activity begins. (A change may also simultaneously
occur in posture, here indeed very broadly with Mr. Nixon rising
from his desk.)
The obvious candidate for illustrations of the Nixon shift
comes from what linguists generally call "code switching," code
here referring to language or dialect. The work of John Gumperz
and his colleagues provides a central source. A crude example
may be cited (Blom and Gumperz 1972:424):
On one occasion, when we, as outsiders, stepped up to a group of
locals engaged in conversation, our arrival caused a significant
alteration in the casual posture of the group. Hands were removed
from pockets and looks changed. Predictably, our remarks elicited
a code switch marked simultaneously by a change in channel cues
(i.e., sentence speed, rhythm, more hesitation pauses, etc.) and by
a shift from (R) [a regional Norwegian dialect] to (B) [an official,
standard form of Norwegian] grammar.
But of course, an outsider isn't essential; the switch can be em-
ployed among the ethnically homogeneous (ibid., p. 425):
Likewise, when residents [in Hemnesberget, northern Norway]
step up to a clerk's desk, greetings and inquiries about family
affairs tend to be exchanged in the dialect, while the business part
of the transaction is carried on in the standard.
Nor need one restrict oneself to the formal, adult world of gov-
ernment and business and its perfunctory service relationships;
the schoolroom will do (ibid., p. 424):
Teachers report that while formal lectures-where interruptions
are not encouraged-are delivered in (B) [an official standard form
of Norwegian], the speaker will shift to (R) [a regional Norwegian
 dialect] when they want to encourage open and free discussion
among students.
By 1976, in unpublished work on a community where Slovene
and German are in active coexistence, matters are getting more
delicate for Gumperz. Scraps of dialogue are collected between
mothers and daughters, sisters and sisters, and code shifting is
found to be present in almost every corner of conversational life.
And Gumperz (1976) makes a stab at identifying what these
shifts mark and how they function:
I. direct or reported speech
2. selection of recipient
3 . interjections
4. repetitions
5. personal directness or involvement
6. new and old information
7. emphasis
8. separation of topic and subject
9. discourse type, e g , lecture and discussion
More important for our purposes here, Gumperz and his cowork-
ers now also begin to look at code-switchinglike behavior that
doesn't involve a code switch at all. Thus, from reconstituted
notes on classroom observations, the Gumperzes provide three
sequential statements by a teacher to a group of first-graders, the
statements printed in listed form to mark the fact that three
different stances were involved: the first a claim on the children's
immediate behavior, the second a review of experiences to comt.,
and the third a side remark to a particular child (Cook-Gumperr.
and Gumperz 1976:B-9):
I. Now listen everybody.
2. At ten o'clock we'll have assembly. We'll all go out together and
go to the auditorium and sit in the first two rows. Mr. Dock, the
principal, is going to speak to us. When he comes in, sit quietly
and listen carefully.
3. Don't wiggle your legs. Pay attention to what I'm saying.
The point being that, without access to bodily orientation and '

tone of voice, it would be easy to run the three segments into a

continuous text and miss the fact that sig
ment of speaker to hearers were occuyring.
7-
-- /
 Forms of Talk

I have illustrated through its changes what will be called

If
footingu2 In rough summary:

I. Participant's alignment, or set, or stance, or posture, or projected

self is somehow at issue.
2. The projection can be held across a strip of behavior that is less
long than a grammatical sentence, or longer, so sentence gram-
mar won't help us all that much, although it seems clear that a
cognitive unit of some kind is involved, minimally, perhaps, a
"phonemic clause." Prosodic, not syntactic, segments are
implied.
3. A continuum must be considered, from gross changes in stance
to the most subtle shifts in tone that can be perceived.
4. For speakers, code switching is usually involved, and if not this
then at least the sound markers that linguists study: pitch, vol-
ume, rhythm, stress, tonal quality.
5. The bracketing of a "higher level" phase or episode of interac-
tion is commonly involved, the new footing having a liminal
role, serving as a buffer between two more substantially sus-
tained episodes.

A change in footing implies a change in the alignment we

take up to ourselves and the others present as expressed in the
way we manage the production or reception of an utterance. A
change in our footing is another way of talking about a change
in our frame for events. This paper is largely concerned with
-
pointing out that participants over the course of their w
cons6ntly change their footing, these c h a n ~ being
a
~h g
a persistent
--_
featG&h&ural talk.
As suggested, change in footing is very commonly language-
linked; if not that, then at least one can claim that the paralinguis-
tic markers of language will figure. Sociolinguists, therefore, can
be looked to for help in the study of footing, including the most
subtle examples. And if they are to compete in this heretofore
literary and psychological area, then presumably they must find
a structural means of doing so. In this paper I want to make a pass
at analyzing the structural underpinnings of changes in footing.
The task will be approached by reexamining the primitive no-
tions of speaker and hearer, and some of our unstated presuppo'
sitions about spoken interaction.
2. An initial statement appears in Goffman (r974:496-559)
 Footing

Traditional analysis of saying and what gets said seems tacitly

committed to the following paradigm: Two and only two in-
dividuals are engaged together in it. During any moment in time,
one will be speaking his own thoughts on a matter and expressing
his own feelings, however circumspectly; the other listening. The
full concern of the person speaking is given over to speaking and
to its reception, the concern of the person listening to what is
being said. The discourse, then, would be the main involvement
of both of them. And, in effect, these two individuals are the only
ones who know who is saying, who is listening, what is being
said, or, indeed, that speaking is going on-all aspects of their
doings being imperceivable by others, that is, "inaccessible."
Over the course of the interaction the roles of speaker and hearer
will be interchanged in support of a statement-reply format, the
acknowledged current-speaking right-the floor-passing back
and forth. Finally, what is going on is said to be conversation or
talk.
The two-person arrangement here described seems in fact to
be fairly common, and a good thing, too, being the one that
informs the underlying imagery we have about face-to-face in-
teraction. And it is an arrangement for which the terms "speaker"
and ."hearer" fully and neatly apply-lay terms here being per-
fectly adequate for all technical needs. Thus, it is felt that without
requiring a basic change in the terms of the analysis, any modifi-
cation of conditions can be handled: additional participants can
be added, the ensemble can be situated in the immediate presence
of nonparticipants, and so forth.
It is my belief that the language that students have drawn on
for talking about speaking and hearing is not well adapted to its
purpose. And I believe this is so both generally and for a consid-
eration of something like footing. It is too gross to provide us
with much of a beginning. It takes global folk categories (like
speaker and hearer) for granted instead of decomposing them into
smaller, analytically coherent elements.
For example, the terms "speaker" and "hearer" imply that
sound alone is at issue, when, in fact, it is obvious that sight is
organizationally very significant too, sometimes even touch. In
 Forms of Talk

the management of turn-taking, in the assessment of reception

through visual back-channel cues, in the paralinguistic function
of gesticulation, in the synchrony of gaze shift, in the provision
of evidence of attention (as in the middle-distance look), in the
assessment of engrossment through evidence of side-involve-
ments and facial expression-in all of these ways it is apparent
that sight is crucial, both for the speaker and for the hearer. For
the effective conduct of talk, speaker and hearer had best be in
a position to watch each other. The fact that telephoning can be
practicable without the visual channel, and that written tran-
scriptions of talk also seem effective, is not to be taken as a sign
that, indeed, conveying words is the only thing that is crucial, but
that reconstruction and transformation are very powerful pro-
cesses.

The easiest improvement on the traditional paradigm for talk is

to recognize that any given moment of it might always be part
of a talk, namely, a substantive, naturally bounded stretch of
interaction comprising all that relevantly goes on from the mo-
ment two (or more) individuals open such dealings between
themselves and continuing un ti1 they finally close this activity
out. The opening will typically be marked by the participants
turning from their several disjointed orientations, moving to-
gether and bodily addressing one another; the closing by tI,eir
departing in some physical way from the prior immediacy of
copresence. Typically, ritual brackets will also be found, such as
greetings and farewells, these establishing and terminating open,
official, joint engagement, that is, ratified participation. In sum-
mary, a "social encounter." Throughout the course of the en-
counter the participants will be obliged to sustain involvement in
what is being said and ensure that no long stretch occurs when
no one (and not more than one) is taking the floor. Thus, at a
given moment no talk may be occurring, and yet the participants -
will still be in a "state of talk." Observe, once one assumes that
an encounter will have features of its own-if only an initiation,
a termination, and a period marked by neither-then it becomes
 Footing

plain that any cross-sectional perspective, any instantaneous slice

focusing on talking, not a talk, necessarily misses important fea-
tures. Certain issues, such as the work done in summonings, the
factor of topicality, the building up of an information state
known to be common to the participants (with consequent
"filling in" of new participants), the role of "preclosings," seem
especially dependent on the question of the unit as a whole.
Giving credit to the autonomy of "a talk" as a unit of activity
in its own right, a domain suigeneris for analysis is a crucial step.
But, of course, only new questions are opened up. For although
it is easy to select for study a stretch of talk that exhibits the
properties of a nicely bounded social encounter (and even easier
to assume that any selected occasion of talk derives from such a
unit), there are apparently lots of moments of talk that cannot be
so located. And there are lots of encounters so intertwined with
other encounters as to weaken the claim of any of them to auton-
omy. So I think one must return to a cross-sectional analysis, to
examining moments of talk, but now bearing in mind that any
broad labeling of what one is looking at-such as "conversation,"
"talk," "discourse"-is very premature. The question of substan-
tive unit is one that will eventually have to be addressed, even
though analysis may have to begin by blithely plucking out a
moment's talk to talk about, and blithely using labels that might
not apply to the whole course of a conversation.

Turn first, then, to the notion of a hearer (or a recipient, or a

listener). The process of auditing what a speaker says and follow-
ing the gist of his remarks-hearing in the communication-sys-
tern sense-is from the start to be distinguished from the social
slot in which this activity usually occurs, namely, official status
as a ratified participant in the encounter. For plainly, we might
not be listening when indeed we have a ratified social place in the
talk, and this in spite of normative expectations on the part of the
speaker. Correspondingly, it is evident that when we are not an
official participant in the encounter, we might still be following
the talk closely, in one of two socially different ways: either we
 Forms of Talk

have purposely engineered this, resulting in "eavesdropping," or

the opportunity has unintentionally and inaZEGiilTy come
..
about, as in "ovqhea~ing."In brief, a ratified w a n t may not
be listening, and osm
-e- he a ratified partici-
-
pant.
Now consider that much of talk takes place in the visual and
aural range of persons who are not ratified participants and whose
access to the encounter, however minimal, is itself perceivable by
the official participants. These adventitious participants are "by-
standers." Their presence should be considered the rule, not the
exception. In some circumstances they can temporarily follow the
talk, or catch bits and pieces of it, all without much effort or
intent, becoming, thus, overhearers. In other circumstances they
may surreptitiously exploit the accessibility they find they have,
thus qualifying as eavesdroppers, here not dissimilar to those
who secretly listen in on conversations electronically. Ordinarily,
however, we bystanders politely disavail ourselves of these latter
opportunities, practicing the situational ethic which obliges us to
warn those who are, that they are, unknowingly accessible, oblig-
ing us also to enact a show of disinterest, and by disattending and
withdrawing ecologically to minimize our actual access to the
talk. (Much of the etiquette of bystanders can be generated from
the basic understanding that they should act so as to maximally
encourage the fiction that they aren't present; in brief, that the
assumptions of the conversational paradigm are being realized.)
But however polite, bystanders will still be able to glean some
information; for example, the language spoken, "who" (whether
in categorical or biographical terms) is in an encounter with
whom, which of the participants is speaker and which are listen-
ers, what the general mood of the conversational circle is, and so
forth. Observe, too, that in managing the accessibility of an en-
counter both its participants and its bystanders will rely heavily
on sight, not sound, providing another reason why our initial
two-party paradigm is inadequate. (Imagine a deaf person by-
standing a conversation; would he not be able to glean considera-
ble social information from what he could see?)
The hearing sustained by our paradigmatic listener turns out
to be an ambiguous act in an additional sense. The ratified hearer
in two-person talk is necessarily also the "addressed" one, that
is, the one to whom the speaker addresses his visual attention and
to whom, incidentally, he expects to turn over the speaking role.
But obviously two-person encounters, however common, are not
the only kind; three or more official participants are often found.
In such cases it will often be feasible for the current speaker to
address his remarks to the circle as a whole, encompassing all his
hearers in his glance, according them something like equal status.
But, more likely, the speaker will, at least during periods of his
talk, address his remarks to one listener, so that among official
hearers one must distinguish the addressed recipient from "unad-
dressed" ones. Observe again that this structurally important
distinction between official recipients is often accomplished ex-
clusively through visual cues, although vocatives are available for
managing it through audible ones.
The relation(s) among speaker, addressed recipient, and
unaddressed recipient(s) are complicated, significant, and not
much explored. An ideal in friendly conversation is that no one
participant serve more frequently, or for a longer summation of
time, in any one of these three roles, than does any other partici-
pant. In practice, such an arrangement is hardly to be found, and
every possible variation is met with. Even when a particular pair
holds the floor for an extended period, the structural implication
can vary; for example, their talk can move to private topics and
increasingly chill the involvement of the remaining participants,
or it can be played out as a display for the encircling hearers-
a miniature version of the arrangement employed in TV talk
shows, or a lawyer's examination of a \ : . i t - - , z x before a jury.
Once the dyadic limits of talk are breached, and one admits
bystanders and/or more than one ratified recipient to the scene,
then "subor ' unication" becomes a recognizable pos-
sibility: t a l k e n e d , timed, and pitched to constitute a
perceivedly limited interference to what might be called the
"dominating communication" in its vicinity. Indeed, there are a
great number of work settings where informal talk is subor-
dinated to the task at hand, the accommodation being not to
another conversation but to the exigencies of work in progress.
Those maintaining subordinate communication relative to a
dominant state of talk may make no effort to conceal that they
are communicating in this selective way, and apparently no
 Forms of Talk

pointed effort to conceal what it is they are communicating. Thus

"byplay":
- subordinated communication of a subset of ratified
participants; "cyssplay": communication between ratified par-
ticipants and bystanders across the boundaries of the dominant
encounter; "sidday": respectfully hushed words exchanged en-
tirely among bystanders. Nature is a pedant; in our culture each
of these three forms of apparently unchallenging communication
is managed through gestural markers that are distinctive and well
standardized, and I assume that other gesture communities have
their own sets of functional equivalents.
When an attempt is made to conceal subordinate communi-
cation, "collusion" occurs, whether within the boundaries of an
encounter (collusive byplay) or across these boundaries (collusive
crossplay) or entirely outside the encounter, as when two by-
standers surreptitiously editorialize on what they are overhearing
(collusive sideplay). Collusion is accomplished variously: by con-
cealing the subordinate communication, by affecting that
the words the excolluded can't hear are innocuous, or by using
allusive words ostensibly meant for all participants, but whose
additional meaning will be caught by only some.
Allied to collusion is "innuendo," whereby a speaker, osten-
sibly directing words to an addressed recipient, overlays his
remarks with a patent but deniable meaning, a meaning that has
a target more so than a recipient, is typically disparaging of it, and
is meant to be caught by the target, whether this be the addressed
recipient or an unaddressed recipient, or even a bystander (Fisher
1076).
A further issue. In recommending earlier that a conversation
could be subordinated to an instrumental task at hand, that is,
fitted in when and where the task allowed, it was assumed that
the participants could desist from their talk at any moment when
the requirements of work gave reason, and presumably return to
it when the current attention requirements of the task made this
palpably feasible. In these circumstances it is imaginable that the
usual ritualization of encounters would be muted, and stretches
of silence would occur of variable length which aren't nicely
definable as either interludes between different encounters or
pauses within an encounter. Under these conditions (and many
others) an "open state of talk" can develop, participants having
 Footing

the right but not the obligation to initiate a little flurry of talk,
then relapse back into silence, all this with no apparent ritual
marking, as though adding but another interchange to a chronic
conversation in progress. Here something must be addressed that
is neither ratified participation nor bystanding, but a peculiar
condition between.
There remains to consider the dynamics of ratified participa-
tion. Plainly, a distinction must be drawn between opening or
closing an encounter, and joining or leaving an ongoing one;
conventional practices are to be found for distinguishably accom-
plishing both. And plainly, two differently manned encounters
can occur under conditions of mutual accessibility, each bystand-
ing the other.3 At point here, however, is another issue: the right
to leave and to join, taken together, imply circumstances in which
participants will shift from one encounter to another. At a
"higher" level, one must also consider the possibility of an en-
counter of four or more participants splitting, and of separate
encounters merging. And it appears that in some microecological
social circumstances these various changes are frequent. Thus, at
table during convivial dinners of eight or so participants, marked
instability of participation is often found. Here a speaker may
feel it necessary to police his listenership, not so much to guard
against eavesdroppers (for, indeed, at table overhearing hardly
needs to be concealed), as to bring back strays and encourage
incipient joiners. In such environments, interruption, pitch rais-
ing and trunk orientation seem to acquire a special function and
significance. (Note how a passenger sitting in the front seat of a
taxi can function as a pivot, now addressing his fellow passengers
in the back seat, now the driver, effectively trusting the driver to
determine whether to act as a nonperson or an addressee, and all
this without the driver's taking his eyes off the road or depending
on the content of the remark to provide participation instruc-
tions.) Another example of structural instability is to be observed
when couples meet. What had been two "withs" provide the
personnel for a momentarily inclusive encounter, which can then
~ . .

3- One standard arrangement is mutual modulation presented as equally

allocating the available sound space; another (as suggested), is differential mut-
ing, whereby those in one of the encounters unilaterally constrain their commu-
nication in deference to the other, or even bring it t o a respectful close.
 Forms of Talk

bifurcate so that each member of one of the entering withs can

personally greet a member of the other with, after which greeting,
partners are exchanged and another pair of greeting interchanges
follows, and after this, a more sustained regrouping can occur.
Consider now that, in dealing with the notion of bystanders,
a shift was tacitly made from the encounter as a point of reference
to something somewhat wider, namely, the "social situation,"
defining this as the full physical arena in which persons present
are in sight and sound of one another. (These persons, in their
aggregate, can be called a "gathering," no implications of any
kind being intended concerning the relationships in which they
might severally stand to one another.) For it turns out that rou-
tinely it is relative to a gathering, not merely to an encounter, that
the interactional facts will have to be considered. Plainly, for
example, speakers will modify how they speak, if not what they
say, by virtue of conducting their talk in visual and aural range
of nonparticipants. Indeed, as Joel Sherzer has suggested, when
reporting on having heard someone say something, we are likely
to feel obliged to make clear whether we heard the words as a
ratified participant to the talk of which they were a part or
whether we overheard them as a bystander.
Perhaps the clearest evidence of the structural significance of
the social situation for talk (and, incidentally, of the limitation of
the conventional model of talk) is to be found in our verbal
behavior when we are by ourselves yet in the immediate presence
of passing strangers. Proscriptive rules of communication oblige
us to desist in use of speech and wordlike, articulated sounds. but
in fact there is a wide variety of circumstances in which we will
audibly address statements to ourselves, blurt out imprecations,
and utter "response cries," such as Oops!, Eek!, and the like (Goff-
man, "Response Cries," 1978 and this volume). These vocaliza-
tions can be shown to have a self-management function,
providing evidence to everyone who can hear that our observable
plight is not something that should be taken to define us. To that
end the volume of the sounding will be adjusted, so that those
in the social situation who can perceive our plight will also hear -
our comment on it. No doubt, then, that we seek some response
from those who can hear us, but not a specific reply. No doubt
the intent is to provide information to everyone in range, but
 Footing

without taking the conversational floor to do so. What is sought

is not hearers but overhearers, albeit intended ones. Plainly, the
substantive natural unit of which self-directed remarks and re-
sponse cries are a part need not be a conversation, whatever else
it might be.
Finally, observe that if one starts with a particular individual
in the act of speaking-a cross-sectional instantaneous view-
one can describe the role or function of all the several members
of the encompassing social gathering from this point of reference
(whether they are ratified participants of the talk or not), couch-
ing the description in the concepts that have been reviewed. The
relation of any one such member to this utterance can be called
his "par-cipation- -.-status" relative to it, and that of all the persons
-CCI
in the gather&&TIie p a r t m n frarnevork" for that moment
of speech. The same two terms can be employed when the point
of reference is shifted from a given particular speaker to some-
thing wider: all the activity in the situation itself. The point of
all this, of course, is that an utterance does not carve up the world .
beyond the speaker into precisely two parts, recipients and non-
recipients, but rather opens up an array of structurally differen-
tiated possibilities, establishing the participation framework in
which the speaker will be guiding his delivery.

I have argued that the notion of hearer or recipient is rather crude.

In so doing, however, I restricted myself to something akin to
ordinary conversation. But conversation is not the only context
of talk. Obviously talk can (in modern society) take the form of
a platform monologue, as in the case of political addresses, stand-
up comedy routines, lectures, dramatic recitations, and poetry
readings. These entertainments involve long stretches of words
coming from a single speaker who has been given a relatively
large set of listeners and exclusive claim to the floor. Talk, after
all, can occur at the town podium, as well as the town pump.
And when talk comes from the podium, what does the hear-
ing is an audience, not a set of fellow conversationalists. Audi-
ences hear in a way special to them. Perhaps in conjunction with
 Forms of Talk

the fact that audience members are further removed physically

from the speaker than a c ~ c ~ n v e r ~ a t i o n might
a l i ~ t be, they have
the right to examine the speaker directly, with an openness that
might be offensive in conversation. And except for those very
special circumstances when, for example, the audience can be
told to rise and repeat the Lord's Prayer, or to donate money to
a cause, actions can only be recommended for later consideration,
not current execution. Indeed, and fundamentally, the role of the
audience is to appreciate remarks made, not to reply in any direct
way. They are to conjure up what a reply might be, but not utter
it; "back-channel" response alone is what is meant to be available
to them. They give the floor but (except during the question
period) rarely get it.
The term "audience" is easily extended to those who hear
talks on the radio or TV, but these hearers are different in obvi-
ous and important ways from those who are live witnesses to it.
Live witnesses are coparticipants in a social occasion, responsive
to all the mutual stimulation that that provides; those who
audit the talk by listening to their set can only vicariously join
the station audience. Further, much radio and TV talk is not
addressed (as ordinary podium talk is) to a massed but visible
grouping off the stage, but to imagined recipients; in fact, broad-
casters are under pressure to style their talk as though it were
addressed to a single listener. Often, then, broadcast talk in-
volves a conversational mode of address, but, of course, merely
a simulated one, the requisite recipients not being there in the
flesh to evoke it. And SG r: brondcL-cttalk may have a "live"
audience and a broadcast audience, the speaker now styling his
projection mainly for the one, now for the other, and only the
music of language can lull us into thinking that the same kind
of recipient entity is involved.
Still further multiplicities of meaning must be addressed.
Podiums are often placed on a stage; this said, it becomes plain
that podiums and their limpets are not the only things one finds
there. Stage actors are found there, too, performing speeches to
one another in character, all arranged so they can be listened in
on by those who are off the stage. We resolutely use one word,
"audience," to refer to those who listen to a political speech and
 Footing

those who watch a play; but again the many ways in which these
two kinds of hearers are in the same position shouldn't blind one
to the very important ways in which their circumstances differ.
A town speaker's words are meant for his audience and are
spoken to them; were a reply to be made, it would have to come
from these listeners, and indeed, as suggested, signs of agreement
and disagreement are often in order. It is presumably because
there are so many persons in an audience that direct queries and
replies must be omitted, or at least postponed to a time when the
speech itself can be considered over. Should a member of the
audience assay to reply in words to something that a speaker in
midspeech says, the latter can elect to answer and, if he knows
what he's about, sustain the reality he is engaged in. But the
words addressed by one character in a play to another (at least
in modern Western dramaturgy) are eternally sealed off from the
audience, belonging entirely to a self-enclosed, make-believe
realm-although the actors who are performing these characters
(and who in a way are also cut off from the dramatic action) might
well appreciate signs of audience a t t e n t i ~ e n e s s . ~
I have suggested that orators and actors provide a ready
contrast to a conversation's speaker, the former having audiences,
the latter fellow conversationalists. But it must be borne in mind
that what goes on upon the platform is only incidentally-not
analytically-talk. Singing can occur there (this being another
way words can be uttered), and doings which don't centrally
involve words at all, such as instrument playing, hat tricks, jug-
gling, and all the other guileful acts that have done a turn in
vaudeville. The various kinds of audiences are not, analytically
speaking, a feature of speech events (to use Hymes's term), but
of stage events.
And from here one can go on to still more difficult cases.
There are, for example, church congregations of the revivalist
type wherein an active interchange is sustained of calls and an-
swers between minister and churchgoers. And there are lots of
4. Maintaining a rigid line between characters and audience is by no
means, of course, the only way to organize dramatic productions, Burmese
traditional theatre providing one example (Becker 1970), our own burlesqued
melodrama almost another.
 Forms of Talk

social arrangements in which a single speaking slot is organiza-

tionally central, and yet neither a stage event with its audience,
nor a conversation with its participants, is taking place. Rather,
something binding is: court trials, auctions, briefing sessions, and
course lectures are examples. Although these podium occasions
of binding talk can often support participants who are fully in the
audience role, they also necessarily support another class of hear-
ers, ones who are more committed by what is said and have more
right to be heard than ordinarily occurs in platform entertain-
ments.
Whether one deals with podium events of the recreational,
congregational, or binding kind, a participation framework spe-
cific to it will be found, and the array of these will be different
from, and additional to, the one generic to conversation. The
participation framework paradigmatic of two-person talk doesn't
tell us very much about participation frameworks as such.

It is claimed that to appreciate how many different kinds of

hearers there are, first one must move from the notion of a con-
versational encounter to the social situation in which the encoun-
ter occurs; and then one must see that, instead of being part of
a conversation, words can be part of a podium occasion where
doings other than talk are often featured, words entering at the
beginning and ending of phases of the program, to announce,
welcome, and thank. This might still incline one to hold that
when words pass among a small number of persons, the
prototypical unit to consider is nevertheless a conversation or a
chat. However, this assumption must be questioned, too.
In canonical talk, the participants seem to share a focus of
cognitive concern-a common subject matter-but less simply so
a common focus of visual attention. The subject of attention is
clear, the object of it less so. Listeners are obliged to avoid staring
directly at the speaker too long 1es.t they violate his territoriality,
and yet they are encouraged to direct their visual attention so as
to obtain gesticulatory cues to his meaning and provide him with
 Footing

evidence that he is being attended. It is as if they were to look

into the speaker's words, which, after all, cannot be seen. It is as
if they must look at the speaker, but not see him.5
But, of course, it is possible for a speaker to direct the visual
attention of his hearers to some passing object-say, a car or a
view-in which case for a moment there will be a sharp difference
between speaker and both cognitive and visual attention. And
the same is true when this focus of both kinds of attention is a
person, as when two individuals talking to each other remark on
a person whom they see asleep or across the street. And so one
must consider another possibility: when a patient shows a physi-
cian where something hurts, or a customer points to where a
try-on shoe pinches, or a tailor demonstrates how the new jacket
fits, the individual who is the object of attention is also a fully
qualified participant. The rub-and now to be considered-is
that in lots of these latter occasions a conversation is not really
the context of the utterance; a physically elaborated, nonlinguis-
tic undertaking is, one in which nonlinguistic events may have
the floor. (Indeed, if language is to be traced back to some primal
scene, better it is traced back to the occasional need of a grunted
signal to help coordinate action in what is already the shared
world of a joint task than to a conversation in and through which
a common subjective universe is generated.6)
One standard nonlinguistic context for utterances is the per-
functory service contact, where a server and client come together
momentarily in a coordinated transaction, often involving money
on one side and goods or services on the other. Another involves
those passing contacts between two strangers wherein the time
is told, the salt is passed, or a narrow, crowded passageway is
negotiated. Although a full-fledged ritual interchange is often
found in these moments, physical transactions of some kind form

5. Overlayed on this general pattern is a very wide range of practices

bearing on the management of interaction. Frequency, duration, and occasion
of mutual and unilateral gaze can mark initiation and termination of turn at talk,
physical distance, emphasis, intimacy, gender, and so forth-and, of course, a
change in footing. See, for example. Argyle and Dean (1965).
6. A useful review of the arguments may be found in Hewes (1973); a
counterview in Falk (1980).
 Forms of Talk

the meaningful context and the relevant unit for analysis; the
words spoken, whether by one participant or two, are an integral
part of a mutually coordinated physical undertaking, not a talk.
Ritual is so often truncated in these settings because it is noncon-
versational work that is being done. It is the execution of this
work, not utterances, that will ordinarily be the chief concern of
the participants. And it is when a hitch occurs in what would
otherwise have been the routine interdigitation of their acts that
a verbal interchange between them is most likely.
A similar picture can be seen in extended service transac-
tions. Take, for example, mother-child pediatric consultations
in Scottish public health clinics, as recently reported by Strong
(1979, esp. chap. 6 ) . Here a mother's business with a doctor
(when she finally gets her turn) is apparently bracketed with
little small talk, very little by way of preplay and postplay, al-
though the child itself may be the recipient of a few ritual
solicitudes. The mother sits before the doctor's desk and briefly
answers such questions as he puts her, waiting patiently, qui-
etly, and attentively between questions. She is on immediate
call, poised to speak, but speaking only when spoken to, almost
as well behaved as the machine that is of so much use to airline
ticketers. The physician, for his part, intersperses his un-
ceremoniously addressed queries with notetaking, note-reading,
thoughtful musings, instruction to students, physical manipula-
tion of the child, verbal exchanges with his nurse and col-
leagues, and movements away from his desk to get at such
things as files and equipment-all af which actioiis ~ ~ p p e war-
ar
ranted by his institutional role if not by the current examina-
tion. The mother's answers will sometimes lead the doctor to
follow up with a next question, but often instead to some other
sort of act on his part. For his social and professional status
allows him to be very businesslike; he is running through the
phases of an examination, or checklist, not a conversation, and
only a scattering of items require a mother's verbal contribu-
tion. And indeed, the mother may not know with any specifi-
city what any of the doctor's acts are leading up to or getting at, -
her being "in on" the instrumentally meaningful sequence of
events in no way being necessary for her contribution to it. So
although she fits her turns at talk, and what she says, to the
 Footing

doctor's questioning (as in the organization of talk), what im-

mediately precedes and what immediately follows these ex-
changes is not a speech environment. What is being sustained,
then, is not a state of talk but a state of inquiry, and it is this
latter to which utterances first must be referred if one is to get
at their organization significance.
O r take the open state of talk that is commonly found in
connection with an extended joint task, as when two mechanics,
separately located around a car, exchange the words required to
diagnose, repair, and check the repairing of an engine fault. An
audio transcription of twenty minutes of such talk might be very
little interpretable even if we know about cars; we would have
to watch what was being done to the car in question. The tape
would contain long stretches with no words, verbal directives
answered only by mechanical sounds, and mechanical sounds
answered by verbal responses. And rarely might the relevant
context of one utterance be another utterance.
So, too, game encounters of the kind, say, that playing bridge
provides, where some of the moves are made with cards, and
some with voiced avowals which have been transformed into
ideal performatives by the rules of the game.
And indeed, in the White House scene presented initially,
the colloquy between Mr. Nixon and Ms. Thomas is not
an embedded part of a wider conversation, but an embedded part
of a ritualized political procedure, the ceremonial signing of a
bill.
One clearly finds, then, that coordinated task activity-not
conversation-is what lots of words are part of. A presumed .,
common interest in effectively pursuing the activity at hand, in
accordance with some sort of overall plan for doing so, is the
contextual matrix which renders many utterances, especially
brief ones, meaningful. And these are not unimportant words; it
i
takes a linguist to overlook them. 1

It is apparent, then, that utterances can be an intimate, func-

tionally integrated part of something that involves other words
only in a peripheral and functionally optional way. A naturally
bounded unit may be implied, but not one that could be called
a speech event.
 Forms of Talk

Beginning with the conversational paradigm, I have tried to de-

compose the global notion of hearer or recipient, and I have
incidentally argued that the notion of a conversational encounter
does not suffice in dealing with the context in which words are
spoken; a social occasion involving a podium may be involved,
or no speech event at all, and, in any case, the whole social
situation, the whole surround, must always be considered. Pro-
vided, thus, has been a lengthy gloss on Hymes's admonition
(1974:54): "The common dyadic model of speaker-hearer spe-
cifies sometimes too many, sometimes too few, sometimes the
wrong participants."
It is necessary now to look at the remaining element of the
conversational paradigm, the notion of speak^.
In canonical talk, one of the two participants moves his lips
up and down to the accompaniment of his own facial (and some-
times bodily) gesticulations, and words can be heard issuing from
the locus of his mouth. His is the sounding box in use, albeit in
some actual cases he can share this physical function with a
loudspeaker system or a telephone. In short, he is the talking
machine, a body engaged in acoustic activity, or, if you will, an
individual active in the role of utterance production. He is func-
tioning as an Animator and recipient are part of the
same level two terms cut from the same
cloth, not social roles in the full sense so much as functional
nodes in 3 con-rmunication system.
But, of course, when one uses the term "speaker," one very
often beclouds the issue, having additional things in mind, this
being one reason why "animator" cannot comfortably be termed
a social role, merely an analytical one.
Sometimes one has in mind that there is an o "-f the
words that are heard, that is, someone who has selected the
sentiments that are being expressed and the words in which they
are encoded.
Sometimes one has in mind that a "grincipal" (in the legalis-
tic sense) is involved, that is, someone whose position is estab-
lished by the words that are spoken, someone whose beliefs have
been told, someone who is committed to what the words say.
 Footing

Note that one deals in this case not so much with a body or mind
as with a person active in some particular social identity or role,
some special capacity as a member of a group, office, category,
relationship, association, or whatever, some socially based source
of self-identification. Often this will mean that the individual
speaks, explicitly or implicitly, in the name of "we," not "I" (but
not for the reasons Queen Victoria or Nixon felt they had), the
"we" including more than the self (Spiegelberg 1973:129-56;
Moerman 1968:153-69). And, of course, the same individual can
rapidly alter the social role in which he is active, even though his
capacity as animator and author remains constant-what in com-
mittee meetings is called "changing hats." (This, indeed, is what
occurs during a considerable amount of code switching, as
Gumperz has amply illustrated.) In thus introducing the name or
capacity in which he speaks, the speaker goes some distance in
establishing a corresponding reciprocal basis of identification for
those to whom this stand-taking is addressed. To a degree, then,
to select the capacity in which we are to be active is to select (or
to attempt to select) the capacity in which the recipients of our
action are present (Weinstein and Deutschberger 1963:454-66).
All of this work is consolidated by naming practices and, in many
languages, through choice among available second-person pro-
nouns.
The notions of animator, author, and principal, taken to-
gether, can be said to tell us about the "productiq~£ormat"
utterance.
,-----..- --
c

When one uses the term "speaker, ' one often implies that
I
of an

the individual who animates is formulating his own text and

staking out his own position through it: animator, author, and
principal are one. What could be more natural? So natural indeed
that I cannot avoid continuing to use the term "speaker" in this
sense, let alone the masculine pronoun as the unmarked singular
form.
But, of course, the implied overlaying of roles has extensive
institutionalized exceptions. Plainly, reciting a fully memorized
text or reading aloud from a prepared script allows us to animate
words we had no hand in formulating, and to express opinions,
beliefs, and sentiments we do not hold. We can openly speak for
someone else and in someone else's words, as we do, say, in
 Forms of Talk

reading a deposition or providing a simultaneous translation of

a speech-the latter an interesting example because so often the
original speaker's words, although ones that person commits
himself to, are ones that someone else wrote for him. As will later
be seen, the tricky problem is that often when we d o engage in
"fresh talk," that is, the extemporaneous, ongoing formulation of
a text under the exigency of immediate response to our current
~ i t u a t i o n it
, ~is not true to say that we always speak our own
words and ourself take the position to which these words attest.
A final consideration. Just as we can listen to a conversation
without being ratified hearers (or be ratified to listen but fail to
do so), so as ratified listeners-participants who don't now have
the floor-we can briefly interject our words and feelings into the
temporal interstices within or between interchanges sustained by
other participants (Goffman 1976275-76, and this volume, pp.
28-29). Moreover, once others tacitly have given us the promise
of floor time to recount a tale or to develop an argument, we may
tolerate or even invite kibitzing, knowing that there is a good
chance that we can listen for a moment without ceasing to be the
speaker, just as others can interrupt for a moment without ceas-
ing to be listeners.

Given an utterance as a starting point of inquiry, I have recom-

mended that our commonsense notions of hearer and speaker are
crude, the firr! 1- ctcntidlj: cr! ~ - - r ? l i ;
~ (, - +>; differentiation of
- , . ; I : '

participation statuses, and the second, complex questions of pro-

duction format.
The delineation of participation framework and production
format provides a structural basis for analyzing changes in foot-
ing. At least it does for the changes in footing described at the
beginning of this paper. But the view that results systematically
simplifies the bearing of participation frameworks and produc-
tion formats on the structure of utterances. Sturdy, sober, socio-
7. David Abercrombie (19652) divides what I here call fresh talk into
conversation, involving a rapid exchange of speaker-hearer roles, and mono-
logue, which involves extended one-person exercises featuring a vaunted style
that approaches the formality of a written form.
 Footing

logical matters are engaged, but the freewheeling, self-referential

character of speech receives no place. The essential fancifulness
of talk is missed. And for these fluidities linguistics, not sociol-
ogy, provides the lead. It is these matters that open up the possi-
bility of finding some structural basis for even the subtlest shifts
in footing.
A beginning can be made by examining the way statements
are constructed, especially in regard to "embedding," a tricky
matter made more so by how easy it is to confuse it with an
analytically quite different idea, the notion of multiple social
roles already considered in connection with "principal."
You hear an individual grunt out an unadorned, naked utter-
ance, hedged and parenthesized with no qualifier or pronoun,
such as:

a directive: Shut the window.

an interrogative: Why here?
a declarative: The rain has started.
a cornmissive: The job will be done by three o'clock.

Commonly the words are heard as representing in some direct

way the current desire, belief, perception, or intention of whoever
animates the utterance. The current self of the person who ani-
mates seems inevitably involved in some way-what might be
called the "addressing self." So, too, deixis in regard to time and
place is commonly involved. One is close here to the expressive
communication we think of as the kind an animal could manage
through the small vocabulary of sound-gestures available to it.
Observe that when such utterances are heard they are still heard
as coming from an individual who not only animates the words
but is active in a particular social capacity, the words taking their
authority from this capacity.
Many, if not most, utterances, however, are not constructed
in this fashion. Rather, as speaker, we represent ourselves
through the offices of a personal pronoun, typically "I," and it is
thus afigure-a figure in a statement-that serves as the agent, a
protagonist in a described scene, a "character" in an anecdote,
someone, after all, who belongs to the world that is spoken about,
not the world in which the speaking occurs. And once this format
is employed, an astonishing flexibility is created.
 Forms of Talk

For one thing, hedges and qualifiers introduced in the form

of performative modal verbs (I "wish," "think," "could," "hope,"
etc.) become possible, introducing some distance between the
figure and its avowal. Indeed, a double distance is produced, for
presumably some part of us unconditionally stands behind our
conditional utterance, else we would have to say something like
"I think that I think. . . ." Thus, when we slip on a word and elect
to further interrupt the flow by interjecting a remedial statement
such as, "Whoops! I got that wrong, . . ." or "I meant to say
.. we are projecting ourselves as animators into the talk. But
.,'I

this is a figure, nonetheless, and not the actual animator; it is

merely a figure that comes closer than most to the individual who
animates its presentation. And, of course, a point about these
apologies for breaks in fluency is that they themselves can be
animated fluently, exhibiting a property markedly different from
the one they refer to, reminding one that howsoever we feel
obliged to describe ourselves, we need not include in this descrip-
tion the capacity and propensity to project such descriptions.
(Indeed, we cannot entirely do so.) When we say, "I can't seem
to talk clearly today," ihat statement can be very clearly said.
When we say, "I'm speechless!", we aren't. (And if we tried to
be cute and say, "I'm speechless-but apparently not enough to
prevent myself from saying that," our description would embody
the cuteness but not refer to it.) In Mead's terms, a "me" that tries
to incorporate its "I" requires another "I" to do so.
Second, as Hockett ( 1 9 6 3 : 1 ~recommends,
) unrestricted dis-
pl-cement in time and place becomes possible, such that our
reference can be to what we did, wanted, thought, etc., at some
distant time and place, when, incidentally, we were active
in a social capacity we may currently no longer enjoy and an
identity we no longer claim. It is perfectly true that when we
say:
I said shut the window
we can mean almost exactly what we would have meant had we
uttered -t he unadorned version:
Shut the window
 Footing

as a repetition of a prior command. But if we happen to be

recounting a tale of something that happened many years ago,
when we were a person we consider we no longer are, then the
"I" in "I said shut the window" is linked to us-the person
present-merely through biographical continuity, something
that much or little can be made of, and nothing more immediate
than that. In which case, two animators can be said to be in-
volved: the one who is physically animating the sounds that are
heard, and an e m k d d e d animator, a figure in a statement who
_L__C:
is present only in a world that is being told about, not in the
world in which the current telling takes place. (Embedded au-
thors and principals are also possible.) Following the same argu-
ment, one can see that by using second or third person in place
of first person we can tell of something someone else said, some-
one present or absent, someone human or mythical. We can
embed an entirely different speaker into our utterance. For it is
as easy to cite what someone else said as to cite oneself. Indeed,
when queried as to precisely what someone said, we can reply
quotatively:
Shut the window
and, although quite unadorned, this statement will be understood
as something someone other than we, the current and actual
animator, said. Presumably, "He (or "she") said" is implied but
not necessarily ~ t a t e d . ~
Once embedding is admitted as a possibility, then it is an
c ~ c ; ?cicp to S P !:-
~ . ' :-- 2 embeddings will be possible, as in
l!:;?i

the following:
To the best of my recollection,
( I ) I think that
(2) I said
( 3 ) I once lived that sort of life.
where (I)reflects something that is currently true of the individ-
ual who animates (the "addressing self"), (2) an embedded
8. Some generative semanticists have argued that any unadorned utter-
ance implies a higher performative verb and a pronoun, e.g., "I say," "aver,"
"demand," etc., the implication being that all statements are made by figures
mentioned or implied, not by living individuals. See, for example, Ross 1970.
 Forms of Talk

animator who is an earlier incarnation of the present speaker and

(3) is a doubly embedded figure, namely, a still earlier incarnation
of an earlier i n ~ a r n a t i o n . ~
Although linguists have provided us with very useful treat-
ments of direct and indirect quotation, they have been less help-
ful in the question of how else, as animators, we can convey
words that are not our own. For example, if someone repeatedly
tells us to shut the window, we can finally respond by repeating
his words in a strident pitch, enacting a satirical version of his
utterance ("say-foring"). In a similar way we can mock an accent
or dialect, projecting a stereotyped figure more in the manner that
stage actors do than in the manner that mere quotation provides.
So, too, without much warning, we can corroborate our own
words with an adage or saying, the understanding being that
fresh talk has momentarily ceased and an anonymous authority
wider and different from ourselves is being suddenly invoked
(Laberge and Sankoff 1979, esp. sec. 3). If these playful projec-
tions are to be thought of in terms of embedding, then stage
acting and recitation must be thought of as forms of embedded
action, too. Interestingly, it seems very much the case that in
socializing infants linguistically, in introducing them to words
and utterances, we from the very beginning teach them to use
talk in this self-dissociated, fanciful way.1°

9. It would be easy to think that "I" had special properties uniquely

bridging between the scene in which the talking occurs and the scene about
.hi(-!- ',crc is talking, for i t refers l?nt!7 to ;. filn-p it? a strttcmcrt 2nd to the
currently present, live individual who is animating the utterance. But that is not
quite so. Second-person pronouns are equally two-faced, referring to figures in
statements and currently present, live individuals engaged in hearing what a
speaker is saying about them. Moreover, both types of pronoun routinely ap-
pear embedded as part of quoted statements:
She said, "I insist you shut the window."
in which case the individual who had served as a live, currently present anima-
tor has herself become a figure in a lower-order statement. The bridging power
of "I" remains, but what is bridged is an embedded speaker to the figure it
describes. The scene in which speaking and hearing is currently and actually --
occurring does not appear except through implicature: the implication that
everyone listening will know who is referred to by "she."
10.In play with a child, a parent tries to ease the child into talk. Using
"we" or "I" or "baby" or a term of endearment or the child's name, and a lisping
sort of baby talk, the parent makes it apparent that it is the child that is being
 It should be clear, then, that the significance of production
format cannot be dealt with unless one faces up to the embedding
function of rnuch~alk.For obviously, when we shift &om saying
\-
somethin.urselves to reporting what someone else said, we are
changing our footing. And so, too, when we shift from reporting
our current feelings, the feelings of the "addressing self," to the
feelings we once had but no longer espouse. (Indeed, a code
switch sometimes functions as a mark of this shift.)
Final points. As suggested, when as speaker we project our-
selves in a current and locally active capacity, then our copartici-
pants in the encounter are the ones who will have their selves
partly determined correspondingly. But in the case of a replay of
a past event, the self we select for ourself can only "altercast" the
other figures in the story, leaving the hearers of the replay undeter-
mined in that regard. They are cast into recipients of a bit of
narrative, and this will be much the same sort of self whomsoever
we people our narrative with, and in whatsoever capacity they
are there active. The statuses "narrator" and "story listener,"
which would seem to be of small significance in terms of the
overall social structure, turn out, then, to be of considerable im-
portance in conversation, for they provide a footing to which a
very wide range of speakers and hearers can briefly shift." (Ad-
talked/&, not to. In addition, there are sure to be play-beings easy to hand-
dolls, teddy bears, and now toy robots-and these the parent will speak for, too.
So even as the child learns to speak, it learns to speak for, learns to speak in
the name of figures that will never be, or at least aren't yet, the self. George
Herbert Mead notwithctanding, the child does not merely learn to rcl'cr to itself
through a name for itself that others had first chosen; it learns just as early to
embed the statements and mannerisms of a zoo-full of beings in its own verbal
behavior. It can be argued that it is just this format that will allow the child in
later years to describe its own past actions which it no longer feels are character-
istic, just as this format will allow the child to use "I" as part of a statement
that is quoted as something someone else said. (One might say that Mead had
the wrong term the child does not acquire a "generalized other" so much as a
u
capacity to embed "particularized others -which others, taken together, form
a heterogeneous, accidental collection, a teething ring for utterances and not a
ball team.) It strikes me, then, that although a parent's baby talk (and the talk
the child first learns) may involve some sort of simplification of syntax and
lexicon, its laminative features are anything but childlike. Nor d o I think parents
should be told about this. A treatment of this issue in another culture is pro-
vided by Schieffelin (1974).
11. One example: A few years ago, the BBC did an hour-length T V
documentary on backstage at the Royal Household. The show purported to
 Forms of Talk

mittedly, if a listener is also a character in the story he is listening

to, as in the millions of mild recriminations communicated be-
tween intimates, then he is likely to have more than a mere
listener's concern with the tale.)
Storytelling, of course, requires the teller to embed in his
own utterances the utterances and actions of the story's charac-
ters. And a full-scale story requires that the speaker remove
himself for the telling's duration from the alignment he would
maintain in ordinary conversational give and take, and for this
period of narration maintain another footing, that of a narrator
whose extended pauses and utterance completions are not to be
understood as signals that he is now ready to give up the floor.
But these changes in footing are by no means the only kind that
occur during storytelling. For during the telling of a tale (as Livia
Polanyi has nicely shown [1977]), the teller is likely to break
narrative frame at strategic junctures: to recap for new listeners;
to provide (in the raconteur's version of direct address) encour-
agement to listeners to wait for the punch line, or gratuitous
characterizations of various protagonists in the tale; or to back-
track a correction for any felt failure to sustain narrative require-
ments such as contextual detail, proper temporal sequencing,
dramatic build-up, and so forth.12
display the Queen in her full domestic round, including shopping and picnick-
ing with her Family. Somehow the producers and stars of the program managed
to get through the whole show without displaying much that could be deemed
inadvertent, revealing, unstaged, or unself-conscious, in part, no doubt, because
much of royal life 1, ;,:L ,. , :, 1 . , < . ' ;ius .i y even in the absence of cameras.
L?
1 '
,L.

But one exception did shine through. The Queen and other members of the
Family occasionally reverted to telling family stories or personal experiences to
their interlocutor. The stories no doubt were carefully selected (as all stories
must be), but in the telling of them the royal personages could not but momen-
tarily slip into the unregal stance of storyteller, allowing their hearers the
momentary (relative) intimacy of story listeners. What could be conceived of
as "humanity" is thus practically inescapable. For there is a democracy implied
in narration; the lowest rank in that activity is not very low by society's stan-
dards-the right and obligation to listen to a story from a person to whom we
need not be in a position to tell one.
I2. Interestingly, the texts that folklorists and sociolinguists provide of
everyday stories often systematically omit the narrative frame breaks that very
likely occurred throughout the actual tellings. Here the student of stories has
tactfully accepted the teller's injunction that the shift in footing required to
introduce a correction or some other out-of-frame comment be omitted from the
official record. Often omitted, too, is any appreciation of the frequency with
 Footing

It was recommended that one can get at the structural basis of

footing by breaking up the primitive notions of hearer and
speaker into more differentiated parts, namely, participation
framework and production format. Then it was suggested that
this picture must itself be complicated by the concept of embed- ,
ding and an understanding of the layering effect that seems to be
an essential outcome of the production process in speaking. But
this complication itself cannot be clearly seen unless one appreci-
ates another aspect of embedding, one that linguistic analysis
hasn't much prepared us for, namely, the sense in which partici-
pation frameworks are subject to transformation. For it turns out
that, in something like the ethological sense, we quite routinely
ritualize participation frameworks; that is, we self-consciously
transplant the participation arrangement that is natural in one
social situation into an interactional environment in which it
isn't. In linguistic terms, we not only embed utterances, we
embed interaction arrangements.
Take collusion, for example. This arrangement may not itself
be common, but common, surely, is apparently unserious collu-
sion broadly played out with winks and elbow nudges in the
obviously open presence of the excolluded. Innuendo is also a
common candidate for playful transformation, the target of the
slight meant to understand that a form is being used unseriously
- a practice sometimes employed to convey an opinion that could
not safely be conveyed through actual innuendo, let alone direct
statement. The shielding of the mouth with the hand, already a
ritualized way of marking a byplay during large meetings, is
brought into small conversational circles to mark a communica-
tion as having the character of an aside but here with no one to
be excluded from it. (I have seen an elderly woman in a quiet
street talking about neighborhood business to the man next door
and then, in termination of the encounter, bisect her mouth with
the five stiff fingers of her right hand and out of one side remark
on how his geraniums were growing, the use of this gesture,

which hearers change footing and inject in passing their own contribution to the
tale (Goodwin 1978, esp. chap. 3 and chap. 4, pt. 5).

153
apparently, marking her appreciation that to play her inquiry
straight would be directly to invoke a shared interest and compe-
tency, not a particularly masculine one, and hence a similarity her
neighbor might be disinclined to confront.) O r witness the way
in which the physical contact, focusing tone, and loving endear-
ments appropriate within the privacy of a courtship encounter
can be performed in fun to an unsuitable candidate as a set piece
to set into the focus of attention of a wider convivial circle. Or,
in the same sort of circle, how we can respond to what a speaker
says to an addressed recipient as though we weren't ratified
coparticipants, but bystanders engaged in irreverent sideplay. Or,
even when two individuals are quite alone together and cannot
possibly be overheard, how one may mark the confidential and
disclosive status of a bit of gossip by switching into a whisper
voice. I think there is no doubt that a considerable amount of
natural conversation is laminated in the manner these illustra-
tions suggest; in any case, conversation is certainly vulnerable to
such lamination. And each increase or decrease in layering-each
u
movement closer to or further from the "literal -carries with it
a change in footing.
Once it is seen that a participation framework can be paren-
thesized and set into an alien environment, it should be evident
that all the participation frameworks earlier described as occur-
ring outside of conversation-that is, arrangements involving an
audience or no official recipient at all-are themselves candidates
for this reframins process; they, too, can be reset into conversa-
;.
,~,,r;l ialk. Anci, oi course, with e a c i ~such crubedding a change

of footing occurs. The private, ruminative self-talk we may em-

ploy among strangers when our circumstances suddenly require
explaining, we can playfully restage in conversation, not so much
projecting the words, but projecting a dumbfounded person pro-
jecting the words. So, too, on such occasions, we can momentarily
affect a podium speech register, or provide a theatrical version
(burlesqued, melodramatic) of an aside. All of which, of course,
provides extra warrant-indeed, perhaps, the main warrant-for
differentiating various participation frameworks in the first place.
It is true, then, that the frameworks in which words are
spoken pass far beyond ordinary conversation. But it is just as
true that these frameworks are brought back into conversation,
 Footing

acted out in a setting which they initially transcended. What

nature divides, talk frivolously embeds, insets, and intermingles.
As dramatists can put any world on their stage, so we can enact
any participation framework and production format in our con-
versation.

I have dealt till now with changes in footing as though the individ-
ual were involved merely in switching from one stance or align-
ment to another. But this image is itself too mechanical and too
easy. It is insufficiently responsive to the way embedding and
ritualization work. For often it seems that when we change voice
-whether to speak for another aspect of ourselves or for some-
one else, or to lighten our discourse with a darted enactment of
some alien interaction arrangement-we are not so much ter-
minating the prior alignment as holding it in abeyance with the
understanding that it will almost immediately be reengaged. So,
too, when we give up the floor in a conversation, thereby taking
up the footing of a recipient (addressed or otherwise), we can be
warranted in expecting to reenter the speaker role on the same
footing from which we left it. As suggested, this is clearly the case
when a narrator allows hearers to "chip in," but such perceivedly
temporary foregoing of one's position is also to be found when
storytelling isn't featured. So it must be allowed that we can hold
the same looting acres, se\r,ial of our turns at talk. And within
one alignment, another can be fully enclosed. In truth, in talk it
seems routine that, while firmly standing on two feet, we jump
up and down on another.
Which should prepare us for those institutional niches in
which a hard-pressed functionary is constrained to routinely sus-
t tain more than one state of talk simultaneously. Thus, through-
out an auction, an auctioneer may intersperse the utterances he
directs to the bidding audience with several streams of out-of-
frame communication-reports on each sale spoken through a
microphone to a recording clerk in another room, instructions to
assistants on the floor, and (less routinely) greetings to friends
and responses to individual buyers who approach with quiet
i
 Forms of Talk

requests for an updating. Nor need there be one dominant state

of talk and the rest styled as asides respectfully inserted at junc-
tures. For example, in a medical research/training facility (as
reported in a forthcoming paper by Tannen and Wallat), a pedia-
trician may find she must continuously switch code, now ad-
dressing her youthful patient in "motherese," now sustaining a
conversa tion-like exchange with the mother, now turning to the
video camera to provide her trainee audience with a running
account couched in the register of medical reporting. Here one
deals with the capacity of different classes of participants to
by-stand the current stream of communication whilst "on hold"
for the attention of the pivotal person to reengage them. And one
deals with the capacity of a dexterous speaker to jump back and
forth, keeping different circles in play.

To end, let us return to the Nixon scene that formed the introduc-
tion to this paper. When Helen Thomas pirouetted for the presi-
dent, she was parenthesizing within her journalistic stance
another stance, that of a woman receiving comments on her ap-
pearance. No doubt the forces at work are sexism and presidents,
but the forces can work in this particular way because of our
general capacity to embed the fleeting enactment of one role in
the more extended performance of another.
When Helen Thomas pirouetted for the president, she was
employing a form of behavior indigenous to the environment of
the ballet, a form that has come, by conventional reframing, to
be a feature of female modeling in fashion shows, and she was
enacting it--of all places-in a news conference. No one present
apparently found this transplantation odd. That is how experi-
ence is laminated.
The news report of this conference itself does not tell us, but
from what is known about Nixon as a performer, a guess would
be that he switched from the high ritual of signing a bill to tKT
joshing of Ms. Thomas not merely as a bracketing device, a signal
that the substantive phase of the ceremony was over, but to show
he was a person of spirit, always capable of the common touch.
 Footing

And, I surmise that although his audience dutifully laughed

loudly, they may have seen his gesture as forced, wooden, and
artificial, separating him from them by a behavioral veil of design
and self-consciousness. All of that would have to be understood
to gain any close sense of what Nixon was projecting, of his
alignment to those present, of his footing. And I believe linguis-
tics provides us with the cues and markers through which such
footings become manifest, helping us to find our way to a struc-
tural basis for analyzing them.
 REFERENCES

Abercrombie, David. 1965. Conversation and spoken prose; Studies in phonetics and
linguistics. London: Oxford University Press.
Argyle, Michael, and Dean, Janet. 1965. "Eye contact, distance and affiliation."
Sociornetry 28:289-304.
Becker, A. L. 1970. "Journey through the night: Notes on Burmese traditional
theatre." T ~Drnrna
P Review 15/3:83-87.
Blom, Jan-Peter, and Gumperz, John J. 1972. "Social meaning in linguistic
structure: Code-switching in Norway." In Direcfonr in S~riolin~uistics,
edited
by John J. Gumperz and Dell H. Hymes. New York: Holt, Rinehart and
Winston.
Cook-Gumperz, Jenny, and Gumperz, John. 1976. "Context in children's
speech." In Papers on Language and Context (Working Paper 46), by Jenny
Cook-Gumperz and John Gumperz. Berkeley: Language Behavior Re-
search Laboratory, University of California, Berkeley.
7he Euening Bulletin (Philadelphia), August 7, 1973.
Falk, Dean, "Language, handedness, and primate brains: Did the Australopithe-
cines sign? American Anthropo/ogist 82 (1980): 78.
Fisher, Lawrence E. 1976. "Dropping remarks and the Barbadian audience."
American Ethnologist 3 /2:22 7-42.
Goffman, Erving. 1974. Frame analysis. New York: Harper and Row.
. 1976. "Replies and responses." Language in Society 5:257-313.
. 1978. "Response cries." Language 54/4.
Goodwin, Marjorie. 1978. "Conversational practices in a peer group of urban
black children." Ph.D. dissertation, Department of Anthropology, Univer-
sity of Pennsylvania, Philadelphia, Pennsylvania.
Gossen, Gary H. 1976. "Verbal dueling in Chamula." In Speech play, edited by
Barbara Kirshenblatt-Gimblett, pp. 121-46. Philadelphia: University of
Pennsylvania Press.
Gumperz, John. 1976. "Social network and language shift." In Papers on Language
and Context (Working Paper 46), by Jenny Cook-Gumperz and John G u m F
erz. Berkeley: Language Behavior Research Laboratory, University of Cali-
fornia, Berkeley.
Hewes, Gordon W. 1973. "Primate communication and the gestural origin of
language." Current Anthropology 14:5-24.
Hockett, Charles. 1963. "The problem of universals in language." In Universals
of language, edited b y Joseph Greenberg, pp. 1-29. Cambridge: MIT Press.
Hymes, Dell H. 1974. Foundations in sociolinguistics. Philadelphia: University of
Pennsylvania Press.
Laberge, Suzanne, and Sankoff, Gillian. 1979. "Anything you can do." In Dis-
course and syntax, edited by Talmy Givon and Charles Li. New York: Aca-
demic Press.
Moerman, Michael. 1968. "Being Lue: Uses and abuses of ethnic identification."
In Essays on the problem of tribe, edited by June Helm, pp. 153-69. Seattle:
University of Washington Press.
Polanyi, Livia. 1977. ''Not so false starts." Working Papers in Sociolinguistics, no. 41.
Austin, Texas: Southwest Educational Development Laboratory.
Ross, John Robert. 1970. "On declarative sentences." In Readings in English trans-
formationalgramrnar, edited by Roderick A. Jacobs and Peter S. Rosenbaum,
pp. 222-72. Waltham, Mass.: Ginn and Company.
Schieffelin, Bambi B. 1979. "Getting it together: An ethnographic approach to
the study of the development of communicative competence." In Studies in
developmental pragmatics, edited by Elinor 0. Keenan. New York: Academic
Press.
Spiegelberg, Herbert. 1973. "On the right to say 'we1: A linguistic and
phenomenological analysis." In Phenomenological sociology, edited by George
Psathas, pp. 129-56. New York: John Wiley and Sons.
Strong, P. M. 1979. 7 l e ceremonial order ofthe clinic. London: Routledge and Kegan
Paul.
Tannen, Deborah, and Wallat, Cynthia. Forthcoming. "A sociolinguistic analy-
sis of multiple demands on the pediatrician in doctor/mother/child in-
teraction."
Weinstein, Eugene, and Deutschberger, Paul. 1963. "Some dimensions of alter-
casting." Sociometry 26/4:454-66.