I came across this linguistics paper today. It talks about the phenomenon where people will talk "through" a pet or preverbal child, either talking as if they are the pet or talking to the pet (usually in high pitched "baby talk") for the purpose of communicating to another person in a way that would be difficult to do directly. My brother informs me that this phenomenon is touched upon in "The Narcissistic Family"; that in some cases the members of a narcissistic family are forced to communicate in this way to circumvent the danger of direct confrontation (or something - I have yet to read the book myself).
I've read about half of it, and it seems interesting. It is a PDF file.
These quotes have a number of errors since a computer program was used to create the "text data". If you read from the PDF file I linked to you won't encounter these "typos".
Based on examples drawn from tape recordings of two middle-class, dual-career
White couples with children 'who audiotaped their own interactions for a week, I-ex
amined how family members mediate interpersonal interaction by speaking as, to, or
about pet dogs who are present in the interaction. Analysis demonstrates that dogs
become resources by which speakers effect a frame shift to a humorous key, buffer
criticism, deliver praise, teach values to a child, resolve potential conflict with a
spouse, and create a family identity that includes the dogs as family members. In this
analysis, I contribute to -an understanding of framing in interaction, including the rel
evance of Bakhtin's (1981) notion of polyvocality for conversational discourse and
demonstrate how family members use pets as resources to mediate their interactions
while constituting and reinforcing their identity as a family.
A couple who live together are having an argument. The man suddenly
turns to their pet dog and says in a high-pitched, baby-talk register, "Mom
my's so mean tonight. You better sit over here and protect me." This makes
the woman laugh--especially because she is a petite 5 ft, 2 in.; her boyfriend
is 6 ft, 4 in. and weighs 285 lb.; and the dog is a lO-lb. Chihuahua mix.
A young woman, the only child of a single mother, is visiting home
from college. At one point her mother tells her, "Pay lots of attention to the
cat; she misses you so much."
These two anecdotal examples illustrate a phenomenon I have been
examining [...]. The phenomenon to which I refer is discursive strategy by which
family members, in communicating with each other, speak through non
verbal third parties-preverbal children or pets. [...] In this study, I focus
exclusively on examples in which the nonverbal third parties are pet
In three of the six examples that follow, family members use pets as
resources in -their interactions by speaking as their pets. I use the term
ventriloquizing to describe the discursive strategy by which a participant
speaks in the voice of a nonverbal third party in the presence of that
To illustrate how I am using the term ventriloquizing, I briefly recap an
example discussed in more detail elsewhere (Tannen, 2003). This example
comes from a third family who participated in our study: Kathy, Sam, and
their daughter Kira who, at the age of 2 years 1 month, was only minimally
verbal. 2 In the following interchange, Kathy was at home with Kira when
Sam returned from work, tired and hungry, and quickly began eating a
snack. Kira, who had eaten dinner earlier with her mother, t..ried to clh~b
onto her father's lap. Sam snapped, "I'm eating!" and Kira began to cry.
Speaking in a high-pitched, sing-song baby talk register, Kathy addressed
Can you say,
-> I was just trying to get some Daddy's attention,
-> and I don't really feel too good, either.
Kathy introduced this utterance by addressing Kiraand asking "Can you
say?" However, by using a baby talk register and the first person singular
"1" (at other times, she animated the child using the second person singular
"we"), she spoke as the child to accomplish a variety of communicative
tasks at once. She (a) indirectly criticized Sam for snapping at their daugh
ter and making her cry, (b) explained Kira's point of view to Sam, and (c)
provided a lesson to Kira that she might one day convey her emotions and
needs more effectively with words rather than with tears. (Elsew~ere,
Kathy states this lesson directly by saying to Kira, "Use your words"-an
injunction that is common to the point of formulaic among preschool
teachers and parents.) The lines indicated by arrows, spoken in the frrst
person and in baby talk register, are those I would characterize as ven
triloquizing because Kathy framed her words as Kira's. She spoke as her
I found the section about other cultures interesting:
Speaking Through Intermediaries: A Cross-Cultural View
Speaking through an intermediary is a phenomenon well documented
in the anthropological literature. Many such examples come from societies
in which it is taboo for individuals in particular kin relations to address
each other directly; others simply illustrate the use of children and pets as
intermediaries in conversational interactions. For example, Schottman
(1993) reports a complex indirect discursive strategy among the Baatombu
of northern Benin by which speakers who harbor grievances assign a prov
erb as a dog's name; they can then invoke the proverb to express their griev
ance simply by calling or addressing the dog. Haviland (1986) examines a
multiparty -interaction in a small Tzotzil-speaking Mexican village to sup
port his claim that conversational mechanisms are designed around multi..
party rather than dialogic interaction. At one point in the interaction, a man
teases an II-year-old boy by playfully offering his daughter as a prospec
tivewife for the boy, adding, "But you mu-st fIrst test her to see if she is any
good" (p. 266). The embarrassed boy does not respond, but his father pro
vides a response for him: "'Am I just a baby that I'll take orders from your
daughter?' you should say that" (p. 268). Haviland observes that on one
hand, the father is teaching his son not to "let such joking remarks pass"
(p. 279), but on the other, the father may also be indirectly communicating
that he himself is not to be taken for a fool.
Another anthropological example comes from Schieffelin's (1990)
study of Kaluli language development in interaction.' Kaluli mothers,
Schieffelin (1990) demonstrates, use the word elema to model for children
what to say.6 In the following example, a 9· month-old baby boy has taken
Bambi's (the author's) net bag. His mother instructs the boy's 2-year-old
sister Meli to chastise her baby brother by saying to Meli, "Don't take~
elema." In response, Meli tells her baby brother, "Don't take-!." The mother
adds, !!Tnis is Bambi~s!---elema. Is it yours?!----elema," and Meli repeats,
"Is it yours?!" (p. 92). The mother then gently takes the bag away from the
Americans might, in a similar situation, expect a mother to speak di
rectly taher baby, instructing him not to take what is .not his. The Kaluli
motl1eris accomplishing the-same result (indeed, she herself physically re
turns the bag to its owner), but she does so in a way that involves her other
child. Thus, as Schieffelin.snows, the Kaluli mother (a) teaches Meli a les
son in values, (b) encourages Meli's language development, and (c) social
izes Meli into the older-sister role toward her brother-a role that, accord
ing to Schieffelin, isflm.damental to Kaluli society.
Another aspect of Kaluli discourse is reminiscent of ventriloquizing.
Kaluli mothers tend to face their babies outward.... Older-children greet and address
infants, and in response to this mothers ... whilernoving them, speak: in a special high
pitched, nasalized register (similar to one that Kaluli use when speaking to dogs.)
These infants look as if they are talking to someone while their mothers speak for
them. (p. 71)
The rest of the paper is dedicated to six examples of this type of communication which are each analyzed. Here's one:
Example 4: Human Interaction as a Resource
for Talking to Pets
The next example· shows Clara using a dog as a resource in interaction
with Neil- and also using Neil as a resource in interaction with the dog.
Here, Clara again uses the watchdog schema to introduce ahumorous key,
this time-to'criticize Neil for a small lapse. This interchange took placedur
ing the evening. Clara is at home, and Neil returns from a brief trip to the lo
cal convenience store. After a few shared observations about a topic of rele
vance to current events, the following exchange occurs:
You leave the door open for any reason?
«short pause, sound of door shutting))
-> <babytalk> Rickie,
-> he's helpin' burglars come in,
-> and you have to defend us Rick.>
Here, Clara speaks to rather than as the dog, yet she uses the same
high-pitched, baby talk register that she uses when she ventriloquizes the
dogs. Rickie can no more understand her discourse than utter it-or ~an
defend h~r and Neil against burglars.
Clara's initial question to Neil ("You leave the door open for any rea
son?") is in itself an indirect linguistic strategy. Rather than tell him di
rectly to close the door, she frames her complaint as a question about -his
motives. That Neil doesn't answer the question is not surprising because
the question is most likely not a literal request for information but an indi..
rect request for action-which Neil apparently provides by closing the
door. (It is not possible to know for sure from the audiotape who closed the
door.) Because the door is heard to be shut before Clara goes on to address
the dog, one may well ask what purpose is served by Clara's utterances to
the dog. On one level, I suspect Clara simply used the dog as a sounding
board for her own inner dialogne. The watchdog schema provides a re
source for verbalizing what would otherwise be unstated, althoughobvi
ous: the reason why it's important to keep the door to the house closed.
I suggest, however, another possible explanation for Clara's utterance
to the dog in this instance. The exchange regarding the door provides a re
source for Clara to talk to the dog much as one taiks nonsense to a baby:
The subject of the talk is not significant, but the sound of the talk, with all
its paralinguistic and prosodic richness, provides an occasion to express the
positive emotion - fondness, attachment - that the speaker feels toward the
child Of, in this case, the dog. At the same time, it does the important work
of including the nonverbal family member in conversational interaction,
initiating, as it were, the child (or the dog) as a family member. In the previ
ous examples, one could say that the dog provided a resource for managing
interpersonal interaction; that level is present in this example as well. How
ever, this example also illustrates yet another function of talking the dog: a
senseinw,hich the interaction among humans:_provides a resource for a pet
owner to express affection and attachment to the dog and to thereby enact
the integration of the petinto the family. Just -as humans reinforce their in
terpersonal connections through talk whether or not there is anything im
portant to talk about at a given moment, humans similarly reinforce their
connections to pets through talk whether or not there is anything to be com..
municated to the pet. Talking to the pet about something that just happened
with a human, therefore, provides material for such talk.
Now, maybe it's just me, but it seems that there might be a passive-aggressive element to Clara's "dog-speak". I suppose it depends on how severe a "mistake" she considers leaving the door open to be. I don't think my machine is clear enough for me to judge these things confidently, but this was my first impression of Clara's comment.
I decided not to post this in "2D Friends" since it's not really "about" the pets, although that's where I initially thought to put it. It probably fits here better.