|
Abstract Index
Conference Index
|
ASD 2000 Conference 17 Abstracts
|
Millennial Dreaming: Washington,
D.C.
|
ABSTRACT
SIGNIFIGANCE OF AUTOMATICALLY DETECTED WORKD RECURRENCES IN
DREAM ASSOCIATIONS
Oral Presentation with UMBERTO BARCARO
The aim of this paper is to discuss the significance of automatically
detected word recurrences in verbal data files including dream reports
and associations with the report items. In fact, assuming that the associations
can provide information about the dream sources, word recurrences often
evidence possibly significant links between different sources.
Each data file consists of: 1) a dream report; 2) the associations
with its various items. The associations were collected asking the subject
about the various items into which the report was divided according to
the concept of activity unit (Foulkes and Schmidt, 1983). Two kinds of
data files were examined: a) data files obtained after forced awakenings
during different sleep conditions (different sleep stages and/or sleep
cycles); control data files in which the associations were with a) sleep
reports provided by another subject; b) literary narrative texts not connected
with dreaming.
A basic assumption underlying the analysis carried out was that the
associations to the various items of a dream report provided information
about the dream sources. Although this assumption may be viewed as controversial
(see Foulkes, 1996), we feel it can be reasonably credited with a certain
degree of validity for two reasons: 1) a hypothesis of this kind, either
explicit (e.g. Cavallero, 1993) or implicit (e.g. Hill et al., 1993) has
contributed to stimulate investigations whose results are consistent with
the recent developments of dream research; 2) dream reports often include
precise circumstantial elements related to the subject's experience which
are immediately identified by the subject's associations: in other words,
a subset of the associations certainly provides reliable information about
the dream sources.
The automatic system consisted of: 1) software procedures implemented
in Microsoft Visual Basic; 2) a Microsoft Access relational database whose
tables associate the words with their roots. The input was either a single
file or a group of files (e.g. those provided by the same subject). The
output was a list of the word root recurrences; the recurrences due to
the obvious contiguity between an item and the associations with it were
discarded; the recurrences of very common words were also discarded. Each
recurrence was accompanied by the reference to the related items of the
report and to its possible significance class.
These indicative significance classes are three:
a) recurrences which appear connected with the logical-narrative structure
of the report;
b) recurrences which are due to the fact that the subject indicated
more than one source for a single item;
c) recurrences which are due to an element common to the sources (different
from each other) of two different items of the report.
Recurrences of classes (b) and (c) indicated some possible links between
the different dream sources. A number of the links evidenced by the automatic
procedure not only escaped the subjects' notice, but were also unexpected
for the analyzer. We feel that attempts of comprehensive dream interpretation
as well as simple hypotheses about the connection between the different
dream sources (e.g. episodes of the subject's life distant in time between
each other) can find a useful tool in this kind of analysis.
NATURE OF INVOLVEMENT WITH DREAMING:
Due to their previous scientific background, the authors are particularly
interested in the study of the psychophysiological aspects of sleep and
in the implementation of automatic methods for dream analysis. At the 1999
Conference at Santa Cruz they presented a communication describing an automatic
procedure for the detection and classification of word root recurrences
in dream reports and the associations with their items. At the next Conference,
they would like to present the development of their study from the point
of view of the significance of the results obtained applying this procedure
to dream reports elicited in different physiological conditions.
UMBERTO
BARCARO, Pisa, Italy
Umberto Barcaro graduated in Physics at Pisa University. He has carried
out research into the quantitative and automatic analysis of electrophysiological
signals. Rosa Calabrese graduated in Medicine at Pisa University.
She specialized in Neurology and in Rehabilitation Medicine. She has carried
out research into the polygraphic study of sleep in pathological conditions.
Corrado Cavallero teaches Research Methods and Data Analysis at the Psychology
Department of Trieste University. Since 1980 he has carried out dream research.
The main topics of his work have been: dream production, technical and
methodological problems in the scoring of dream texts; the use of free
associative technique as an experimental paradigm in dream research.
Roberta Diciotti obtained a Diploma as a Program Analyst at the University
of Siena. She is a technical collaborator at the Istituto di Elaborazione
della Informazione of the National Research Council, where she works at
the Data Base Management System. Carlo Navona graduated in Computer
Science at Pisa University. He has carried out research on the automatic
analysis of human EEG, in particular spontaneous EEG, transient evoked
potentials and steady-state evoked potentials.
Umberto Barcaro was born at Rovigo, Italy, in 1946. He graduated in
Physics in 1968 at Pisa University. He is now Associate Professor at the
Computer Science Department of Pisa University and research collaborator
with the National Research Council at Pisa.
Contact information:
Umberto Barcaro
Pisa, Italy
Email:barcaro@iei.pi.cnr.it |
