LINGUIST List 6.607

Tue 25 Apr 1995

Sum: Spontaneous Speech

Editor for this issue: <>


  1. , Sum: 'spontaneous' speech

Message 1: Sum: 'spontaneous' speech

Date: Mon, 24 Apr 1995 09:58:24 Sum: 'spontaneous' speech
From: <>
Subject: Sum: 'spontaneous' speech

Content-Length: 9197

Two months ago, I've posted a query on the List about possible definitions of
the term 'spontaneous speech'.
Although I replied within weeks separately to each of you who answered, I
still have to keep my promise and post a comprehensive summary of this
'electronic poll'.
''Dear colleagues,
I'm conducting a poll about definitions of 'spontaneous speech' in
This term is now currently used in Phonetics, and it has been applied to
Sociolinguistics (Labov: Sociolinguistic Patterns, 1972) and
Psycholinguistics (Goldman-Eisler: Psycholinguistics: Experiments in
Spontaneous Speech, 1968)
I would be grateful to anyone who can answer the following questions, and
make comments related to the matter:
1. What does 'spontaneous speech' mean for you? Please give a short
2. Do you use this term in your research or academic work? Why?
3. Any comments?
Your help would be a valuable contribution to my Ph.D.
Please reply directly to my address:
If any interest, I'll post a summary about the results of the poll.
15 people form from 7 different countries answered. The answers came from:
USA - 6, GB - 3, Holland - 2, Sweden - 1, Norway - 1, Singapore - 1, Fiji
Islands - 1). People answered were from 5 different backgrounds, such as:
Linguistics (6) no particular field specified, Phonetics (3),
Sociolinguistics (2), Automatic Speech Recognition (2), and Psycholinguistics
Thank you for the valuable remarks, comments and suggestions for further
references of people I mention in the same order I've got their e-mails:
Frances Ingemann, Mary Howe, Hilde Hasselgerd, Maurice Wong, Charles Scott,
Russell J. Collingham, Rianne Doeleman, Jim Milroy, Knud Lambrecht, Jacob
Dempsey, Jan Tent, Max Wheeler, Helmer Strik, Anthea F. Gupta, Duncan
The following studieshave been indicated:
* Nils Erik Enkvist (1982), Impromptu speech, structure and process. in:
Impromptu Speech: A Symposium. Abo, Abo Akademi, Series Publications if the
Research Institute of Abo Akademi Foundation.
* Chao Yuen Ren (1968), Types of Discourse. in: A grammar of spoken Chinese,
pp.17-18, section 1.2.3.)
*L. Milroy, Observing and Analyzing Natural Language (no further reference
The reason why I've posted this question is the following:
I'm a linguist-phonetician working on naturally occurring (let's call them
'spontaneous') media speech in French. I'm interested in comparing rhythmic
features in a French writer's media speech recorded in different speech
situations (interview, debate, reading-aloud..etc.), and during her 30 years
of career. While I've tried to define the corpus I'm dealing with in the most
exact possible terms, I've realized that the meaning of 'spontaneity' seemed
less easy 'to grasp' than I used to think. Different fields seemed also
focusing on different aspects. Given the fact that the term is widely
accepted and used (see answers on Q2), but I still haven't seen any reviews
about works done in this matter in different fields (if you did please let me
know), I've decided to post this query.
Answers to the first question indicate that there is not only one, but at
least four different ways of dealing with naturally occurring speech data in
Linguistics in general. The definition chosen depends on the field and the
particular study undertaken. 'Spontaneous speech' is a
(1) type or 'mode' of speech production opposed to 'read-aloud' speech;
(2) real-time generated, unplanned and non-rehearsed type of encoding
linguistic information;
(3) casual 'way of speaking' or 'style', characterizing informal speech
(4) naturally occurring, non-experimental type of speech event of any kind.
NON-READ. This seems to be the most specific definition. 9 out of 15 people
mentioned it: phoneticians (3), engineers in Automatic Speech Recognition and
Understanding (2), sociolinguist (1), and linguists (3) no field specified.
Everything which in not read-aloud is considered 'spontaneous'. 'Spontaneity'
is defined as a 'way' or 'mode' of encoding that is different from
'oralizing' a written code. Everything encoded from memory (but not in the
sense of rehearsed) is 'spontaneous': "not read", "definitely not scripted",
"not reading style"...etc. "Answers to questions are considered 'spontaneous'
as long as no answers are provided for the speakers to read [...] the
Macrophone database sponsored by the Linguistic Data Consortium makes the
distinction between read responses and spontaneous responses in this manner"
(M.Wong). People also noticed that this 'spoken mode' of speech production is
defined negatively as compared to reading aloud (M. Wheeler)
NON-REHEARSED. Specified in 7 answers in the following terms: "relatively
unplanned", "unplanned", non "monitored", "unprepared", "non reciting
memorized or rehearsed", not 'pre-determined". The focus is clearly on the
internal processes of speech encoding. The more this encoding deals with less
well-integrated elements in real-time, the speech is considered
'spontaneous'. One might recall extended research done about these questions
in Psycholinguistics, Goldman-Eisler...etc. An important aspect of this
definition is: in what extent a message can be considered planned or
unplanned? "I would mean 'unplanned speech' in contrast to any discourse that
I had some opportunity to consider in advance of uttering. But actually, now
that I think about it, it seems highly unlikely that one's ENTIRE 'planned'
discourse would be really so...". (C. Scott).
CASUAL STYLE. This way of considering 'spontaneity' in speech processing is
related to that of the so-called 'vernacular' or informal speaking 'style' in
Sociolinguistics (see Labov for discussion). Four people (2 sociolinguists
and 2 linguists no field specified) mentioned it. While commenting this
definition of 'spontaneity' in speech, J. Milroy, among others, points out
that: "I'm inclined to think that if it's not in laboratory conditions and
not strictly controlled by the interviewer (questions requiring relevant
replies) all the speech that we collect by field methods is spontaneous, EVEN
WHEN IT IS CAREFUL AND FORMAL [emphasis by me]."
NATURALLY OCCURRING. Four people expressed the most general and wide
acception of the term 'spontaneous' speech, among them 1 phonetician, 1
psycholinguist and 2 linguists. For them, as for J. Milroy above, everything
produced in a relevant, really existing speech situation is 'spontaneous':
"that would have occurred even if my tape recorder had not been on." (M.
Howe). As for speech elicited in experiments or even just recorded, 2 people
suggested the term 'semi-spontaneous' (R. Doeleman) or 'peudo- or
quasi-spontaneous' (D. Markham). Phoneticians joining the discussion pointed
out that popular methods of eliciting so-called 'spontaneous' speech form
subjects in, for ex.,'picture descriptions' can't really meet this criterion,
because "you shouldn't instruct (and therefore, restrict) subjects but
instead let them say something SPONTANEOUSLY" (H.Strik). What 'really
spontaneous' is: "where the informant has not been instructed to speak IN A
CERTAIN WAY" (D. Markham). And F. Ingeman to joint them when she says: "I do
not put a lot of trust in data given in response to direct elicitation of
grammatical information that is not supported by spontaneous speech".
Answers given to the second question indicate that 14 out of 15 people
answered use currently the term 'spontaneous speech' in their academic work
or teaching. Three of them indirectly, one of them directly specified why:
"This is a very important concept [registers] in French syntax" (K.
Lambrecht); "This distinction [read-spontaneous] is important in spoken
language understanding systems (M. Wong); "Spontaneous speech is THE MOST
interesting type of linguistic behavior in my book." (J. Tent).
Any comments have been relevant and interesting for me. They also reflected
people's interest in the subject. Actually, the number of answers to this
particular question (15) is higher than the average of responses collected in
other summaries on the List during this period (from February 4 to 28). (Of
course, questions related to the O.J. Simpson trial or IBM TV commercial
advertisements excepted...) Thanks again for your help.
Susan Fagyal
Institut de Phonetique de Paris
Research Institute for Linguistics Hungary
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue