Transcript Everyday Terms: How We Seem to Use Them

Toward a Selectionist Ontology in
Behavior Analysis
Ailun – Lecture 3
S. Glenn - AILUN 2008
Part I: Ontology
• All scientific domains deal with ontological
questions regarding “what exists”
• “There are only two bona fide ontological
categories: individuals and classes”1
• Ontological status of theoretical entities is
theory dependent2
• Ontological issues herein derive from a
selectionist theoretical approach
S. Glenn - AILUN 2008
Ordinary Language Classification
Classes
and Subclasses
Individuals (and their Parts)
Astronomical bodies
Stars
Planets
Sun, Earth
Sun (core)
Earth (Himalayas)
Hurricanes
Category 3
Category 4
Hurricane Rita, Hurricane Bart
Rita (eye of Hurricane Rita)
Bart (the eye wall of Hurricane Bart)
Lever pressing operants
Human lever pressing
Rat lever pressing
Tom’s pressing, Rat 101’s pressing
Tom’s lever pressing (Press #1)
R 101’s lever pressing (Press #2000)
Corporations
For-profit corporation
Not-for-profit corporation
Fiat, Assoc for Behavior Analysis (ABA)
Fiat, Inc (Lingotto factory)
ABA (Exec Council)
S. Glenn - AILUN 2008
Characteristics of Individuals1,2
• Spatiotemporally restricted (e.g. Earth)
– Have duration (extend in time)
– Have beginning and end
– Are located in space
– Can be given a proper name
• Can have parts (e.g. Himalayas)
• Can be parts of a larger whole (Our solar system)
• Can change over time (e.g. newborn becomes adult)
• Can relate as a unitary object/event to other
objects/events (Earth revolves around Sun)
S. Glenn - AILUN 2008
Characteristics of Classes1,2
• Defined intensionally by their common
properties
• Don’t have parts, have members (which are
individuals)
• Members belong to a class because they have
the characteristics of the class
• Spatiotemporally unrestricted (the class isn’t
located at any particular segment of
space/time)
• A class doesn’t change, develop or evolve
• Not limited to finite number of members
S. Glenn - AILUN 2008
Individuals and Classes in Science3
• Individuals are what scientists observe, measure and classify, and what are
“acted upon by natural processes” (p. 147)
• A goal of all sciences is “to discover ways to divide up the world into classes
that function in natural regularities” (p. 147) Such classes are called “natural
kinds”
• The ‘natural regularities’ are repeated observed relations among individuals
constituting a ‘natural kind’
• The repeated relations (natural regularities) are described by scientific
generalizations (principles) that specify the processes accounting for change
in the individuals
• The terms in the principles are class terms; they do not refer to individuals
• The principles “explain” the observed changes in individuals by appealing to
the regularities specified in the scientific generalizations
• The classes specified in principles are, over time, organized in a framework
that allows prediction and control of the observed natural or social
phemonena
S. Glenn - AILUN 2008
Part II: Individuals & Classes
in Operant and Organic Selectionist Theory
Classes
Individuals
H. Sapiens
C. livia
B. F. Skinner
S 301
Fred’s
Gene #55
301’s
Gene #22
Genes
Organisms
Species
S. Glenn - AILUN 2008
Units of Selection at Two Levels
Responses : Operants
::
Organisms : Species6
All the above terms specify both individuals and classes
• Individuals
•
– Specific responses (Tom’s 1st lever press) are parts of specific operants
(Tom’s lever pressing)
– Specific organisms (Tom) are parts of specific species (H.Sapiens)
Classes
– Tom’s lever presses are members of the class “responses” just as Tom is
a member of the class “organisms”
– Tom’s lever pressing operant is a member of a class called “operants”
just as Tom is a member of classes called “organisms”
• The principles of selectionist theories specify relations among the
classes
– Evolutionary theory is about organisms, genes, and species, not about
Tom, gene # 40 or H. sapiens
– Operant theory is about responses, operants (and perhaps at some
time, firing patters in NS). Not about Tom’s 1st leverpress, or his history
of leverpressing
S. Glenn - AILUN 2008
Operants as Ontological Individuals5
• Consider each behavioral event (localized with respect to
organism, time, place) as an individual having specifiable
properties (form, IRT, duration, force, effect)
• Some behavioral events are reliably followed by specific
changes in the environment (they are part of a
response/consequence contingency)
• If a contingency between behavioral events and specific
consequences results in an increase in the frequency of
behavioral events having those properties, a lineage of
‘ancestor/descendant’ behavioral events forms
• This lineage is an individual – its properties (rate, variability,
average IRT, topography) can change over time
• The individual responses are parts of that particular
individual operant lineage
• Origin of the lineage, its properties, and the properties of
its parts (behavioral events) are caused by (a function of)
the contingency (recurring response/consequence
relations) that has operated
in the past
S. Glenn - AILUN 2008
Operant as a“Natural Kind”5
A class term in a generalized principle of behavioral selection
• The process of behavioral selection accounts for origin and
maintenance of operant lineages and the characteristics of
their parts (responses)
• Each operant lineage is an individual that changes over time
as the properties of its parts are differentially selected by
their relation to their external environment
• The class of lineages that are formed and altered in this way
is a class of “operants”.
• This class (“operants”) plays a fundamental role in a theory
of behavior change (learning).
• The category of behavioral events that exist as parts of
lineages formed by contingencies of reinforcement is the
category of “operant behavior”.
S. Glenn - AILUN 2008
Confusion in Operant Ontology
• Operant: “A functional category of behavior”
– Does this mean that the category or the behavior is
functional?
– If the behavior is what is functional, then this means
“operant” is a category of behavior defined in terms of
function (alters environment, produces consequences)
– If the category is what is functional, then this means
“operant” is a category of behavior that functions in a
specific way in scientific principles
• Could mean either or both, but important to recognize the
difference
S. Glenn - AILUN 2008
Confusion in Operant Ontology
• An operant: “A class of acts all of which have the same
environmental effect”, e.g., in lab, “all acts that have the
effect of depressing the lever”)6
– If “all acts” having the “same effect” are limited to those of a
specific organism,
– then “operant” suggests a spatio-temporal locus so this lever
pressing is an individual, not a class.
– Another possible meaning: all acts of any organism that have the
effect of depressing a lever (spatiotemporlly unrestricted)
– Then a genuine class—but not likely to function in any scientific
laws
• Leverpresses of a rat for food and lever presses of the same rat
for water would belong to different operants
• Responses having the same effect (switch closure) but different
consequences are different operants
S. Glenn - AILUN 2008
Confusion in Operant Ontology
Catania
• Operant: “A class of responses”
– “the class is defined in terms of [specific] environmental
effect” = descriptive definition
– “a class modifiable by the consequences of responses in it
[the class] = functional usage
– Note: By the second definition, the class must be modifiable
(able to change over time)
• Classes don’t change over time
• What actually is modified?
– Note: The “function” in the first definition is the function of
the responses in the class
– The “function” in the second definition is the function of the
consequences for the class
S. Glenn - AILUN 2008
A Clearer Ontology8
Donahoe and Palmer
• “Operants: responses that are less reliably evoked by stimuli
than respondents, or for which the stimulus is not well
specified” (p. 360)
• Similar to Skinner’s “uncommitted” behavior (movement)
• Perhaps not easily identified as “responses”
• “ ’Operant’ signifies that the response operates on the
environment to produce the eliciting stimulus [consequence],
and that the stimulus guiding the response may not be well
specified” (p. 38)
• Note: Both these definitions seem to pertain to the operant
category and avoid using class terminology in defining
spatiotemporally localized phenomena
S. Glenn - AILUN 2008
Summary: Ontology in Behavior Analysis
• Operant as individual (e.g. Tom’s leverpresses)
– An individual localized with respect to Tom’s lifetime (“seen” as a
whole in the curves of a cumulative record)
– Made up of more localized individuals: individual presses (responses)
are parts of the whole (seen in each up-tick of the record)
– The whole operant evolves as the characteristics of its parts change as
a result of selection processes occurring at the behavioral level
• Operant as a natural kind (a class having function in natural
regularities described by scientific principles)
– Class members: Tom’s individual leverpressing operant and all other
individual operants of all organisms, living or dead or yet to be
– Behavior analysis principles describe relations between the class
called “operants” and other classes (e.g. reinforcers, discriminative
stimuli, motivating operations, conditional stimuli, etc.)
– The principles can be used to understand, predict and change the
course of individual operants
S. Glenn - AILUN 2008
References
1 Ghiselin, M. T. (1999). Metaphysics and the Origin of Species. Albany: State
University of New York Press (p.37)
2 Hull, D. L. (1977). The ontological status of species as evolutionary units. In R. Butts
& J. Hintikka (Eds). Foundational Problems in Special Sciences (pp. 91-102).
Dordrecht-Holland: D. Reidel Publishing Co. Reprinted in D. L. Hull (1989), The
Metaphysics of Evolution (pp.70-88). Albany, NY: State University of New York
Press.
3 Hull, D. L. (1981). Units of evolution: A metaphysical essay. In U. L. Jensen & R. Harre
(Eds). The Philosophy of Evolution. Brighton: Harvester Press. Rerpinted in R.N.
Brandon & R. M. Burian (1984). Genes, Organisms, Populations: Controversies over
the units of Selection (pp. 142-160. Cambridge, MA: MIT Press.
4 Skinner (1953). Science and Human Behavior, New York: Free Press
5 Glenn, S. S., Ellis, J., & Greenspoon, J. (1992). On the revolutionary nature of the
operant as a unit of behavioral selection. American Psychologist, 47, 1329-1336.
6 Baum, W. M. Understanding Behaviorism: Science, Behavior, and Culture. Ner York:
HarperCollins College Publishers. (Quotation p. 75)
7 Catania, A. C. (2007). Learning (4th Edition). Cornwall-on-Hudson: Sloan Publishing.
(Quotations from Glossary)
8 Donahoe, J. W. & Palmer, D.C. (1994). Learning and Complex Behavior. Boston:
Allyn and Bacon.
S. Glenn - AILUN 2008