-
William P. Fisher, Jr.
Research Associate, BEAR CenterUniversity of California,
Berkeley, CA, USA
&Living Capital Metrics LLC, Sausalito, CA, USA
BEAR Seminar28 January 2020
The Semiotics of Identifiable Models in the Economy of
Thought:
Making Improved Measurement More Widely Available in Psychology
and the Social Sciences
-
A simple problem
-
Outline
• Some definitions• Semiotics
• Identifiability
• Economy of thought
• Philosophical and historical considerations
• A plan for action
-
Semiotics
• Dictionary definition:• The study of signs and symbols
• The study of their use or interpretation
• Semiotic triangle:• A concept: ‘chair’
• A word: “chair”
• A thing: chair
Idea
WordThing
-
Identified and unidentified models
• Originated in Frisch’s concept of autonomy of data from
econometric model (Aldrich, 1989)
• "A system of autonomous equations has the property 'that it is
possible that the parameters in any one of the equations could in
fact change ... without any change taking place in any of the
parameters of the other equations.‘” (Girshick & Haavelmo,
1947, p. 106)
• Aldrich (1989, p. 15) “Autonomy is significant because an
equation (or its parameters) in a system of autonomous relations
will be ‘detachable’ or ‘projectible’. Detachability matters both
for interpretation and for ‘policy simulation.’”
-
Economy of thought
• Mach (emphasis added)• "It is the object of science to
replace, or save, experiences, by the
reproduction and anticipation of facts in thought. Memory is
handier than experience, and often answers the same purpose. This
economical office of science, which fills its whole life, is
apparent at first glance...”
• "Language, the instrument of this communication, is itself an
economical contrivance.”
Mach, E. (1883/1919). The science of mechanics: A critical and
historical account of its development (T. J. McCormack, Trans.)
(4th ed.). Chicago: The Open Court Publishing Co., p. 481.
-
Science and language anticipate new facts by structuring
perception and focusing attention
-
chair
椅子la sillaкрісло
كرسي
No abstract concept is ever fully realized in any actual
reference to any concrete thing.
-
The economy of language works because , generally speaking,
linguistic models are identified.
a) the formal conceptual model of the ‘chair’ is invariantly
represented by the standardized arbitrary abstraction of the word
“chair” in relation to all possible actual chairs;
b) standardized pronunciation and inscription of the word
“chair” instrumentally mediates the relation between the formal
ideal and any actual chairs being referred to; and
c) actual chairs as places to sit can encompass a wide range
from metaphorical places on the ground, rocks, or logs; to
functional kitchen and desk chairs; to comfortable chairs; to
dysfunctional decorative or elaborate ceremonial chairs; to
positions of rank, as endowed or committee chairs.
-
‘Chair’
Chair “Chair”
So language gives us cognitive models that work because the
models are identified.
Learning a language by imitating established usage is easier
than inventing our own languages, and then translating between ours
and everyone else’s.
Both poetry and science are labor-saving devices that bring
empirically repeatable and formally explained effects into
language.
-
‘Meter’Theory
Data Instrument
Meter “Meter”
Scientific models are identified.
Length of the path travelled by light in vacuum during a time
interval of 1/299792458 of a second
-
Autonomy and identifiability flip descriptive statistical
modeling into prescriptive scientific modeling
• Frisch “realised that a model was not a correctly specified
structural model just because it fitted well.”
• Before the introduction of autonomy as a criterion, data
analysis was the driver. The goal was to “first find relations in
the data and then make sense of them. Here [with autonomy] the
sequence was reversed: define the relations that made economic
sense and estimate them.”
Aldrich, J. (1989). Autonomy. Oxford Economic Papers, 41, p. 25.
(emphasis added)
-
Anticipatory model identifiability in philosophy
• "...reason has insight only into that which it produces after
a plan of its own.”
• Reason “must itself show the way with principles of judgment
based upon fixed laws.”
• “Accidental observations, made in obedience to no previously
thought-out plan, can never be made to yield a necessary law, which
alone reason is concerned to discover. “
• Kant, I. (1929/1965). Critique of pure reason (N. K. Smith,
Trans.) (Unabridged). New York: St. Martin's Press, p. 20.
-
• Whitehead (emphasis added)
• “Symbolization of the associative law (x+y=y+x), for instance,
simplifies the representation of the idea that 'If a second number
be added to any given number the result is the same as if the first
given number had been added to the second number'. This example
shows that, by the aid of symbolism, we can make transitions in
reasoning almost mechanically by the eye, which otherwise would
call into play the higher faculties of the brain.”
Whitehead, A. N. (1911). An introduction to mathematics. New
York:
Henry Holt and Co., p. 61.
Representing semiotically stable relationships in an economy of
thought
-
Economy of thought supports learning as habituated behaviors•
“…the very heart, blood and sinews of learning is creation of
habitudes.”
• “Habit does not preclude the use of thought, but it determines
the channels within which it operates. Thinking is secreted in the
interstices of habits.“
Dewey, J. (1954). The public and its problems. Athens, Ohio:
Swallow Press, Ohio University Press
-
Autonomy and identifiability in history of science
• “…to Copernicus' mind the question was not one of truth or
falsity, not, does the earth move? He simply included the earth in
the question which Ptolemy had asked with reference to the
celestial bodies alone; what motions should we attribute to the
earth in order to obtain the simplest and most harmonious geometry
of the heavens that will accord with the facts?“
• Burtt, E. A. (1954/1932). The metaphysical foundations of
modern physical science [First edition published in 1924] (Rev.
ed.). Garden City, New York: Doubleday Anchor, p. 39.
-
Conceptual idealization in history of science
• The "law of inertia is not the kind of thing you would
discover by mere photographic methods of observation--it required a
different kind of thinking-cap, a transposition in the mind of the
scientist himself; for we do not actually see ordinary objects
continuing their rectilinear motion in that kind of empty
space....“
• Butterfield, H. (1957). The origins of modern science (revised
edition).New York: The Free Press, pp. 16-17. (emphasis added)
-
Unrealistic conceptual ideals in science
• Newton's first law "...speaks of a body...which is left to
itself. Where do we find it? There is no such body. There is also
no experiment which could ever bring such a body to direct
perception. But modern science…is supposed to be based upon
experience. Instead, is has such a law at its apex. This law speaks
of a thing that does not exist. It demands a fundamental
representation of things which contradict the ordinary.“
• Heidegger, M. (1967). What is a thing? (W. B. Barton, Jr.
& V. Deutsch, Trans.). South Bend, Indiana: Regnery/Gateway, p.
89.
-
Rasch’s involvement in early development of the concept of
identified models
• Autonomy redefined as identifiability by Koopmans and Reiersol
(1950)• In a footnote, thank Rasch and Thurstone for “fruitful”
discussions
in 1947 at Cowles Commission meeting at University of
Chicago
• Rasch (1953, p. 65) also refers to these conversations as
“fruitful”
• Rasch (1977): “...the Models for Measuring are an elaboration
in the same direction of the class of distributions called the
Darmois-Koopmans exponential family, and discussed extensively
since 1935.”
-
Reversing the sequence from descriptive to prescriptive, from
empiricist to rationalist• Identified models have a uniquely
determined statistical meaning.
• Frisch’s work with autonomy and identifiability explain his
astonishment in 1959 at Rasch’s “disappearing parameter.”
• “Haavelmo gave three reasons for being concerned with
autonomy: autonomous relations are likely to be more stable; they
are more intelligible; they are useful for policy analysis”
(Aldrich, 1989, p. 27)
• "'The principal task of economic theory is to establish such
relations as might be expected to possess as high a degree of
autonomy as possible‘” (Aldrich 1989, p. 28, quoting Haavelmo,
1943b, p. 29).
-
Beyond centralized analysis to distributed multilevel
implementations of instruments calibrated in consensus standard
unit
• Different but equivalent models should support the same
substantive inferences concerning comparisons within and across
ability and difficulty distributions.
• But the primary reason for valuing identifiable models is not
analytic.
• Facilitating new economies of thought is possible only by
means of well formed semiotic relationships in identified
models.
-
Beyond centralized analysis to distributed multilevel
implementations of instruments calibrated in shared consensus
standard unit
• Semiotically stable identified models support the cultivation
of multilevel knowledge infrastructures.
• So how can new understandings built into successful models be
packaged in portable calibrated instruments and distributed to end
users in reliable, quality assured forms?
-
Discourse on method to date exhibits reactive shifts between
extreme positions of positivist objectification and radical
contextualism.
• Modern positivist objectification• Embraces universalized
generalization of one set of reified data.• The facts to be
emphasized in any given situation are the ones that meet
immediate
political and economic ends, satisfying the interests of a
particular agenda.
• Postmodern radical contextualism• Embraces incommensurability
of fragmented local values.• The facts emphasized in any given
situation emerge as a function of interests that
have to be acknowledged and constrained to serve the greater
good.
• Unmodern/amodern integrated methodological pluralism• Finds
general in specific and vice versa• The facts to be emphasized in
any given situation are contextualized within formal
conceptual determinations embodied in standardized media and
units with known uncertainties.
-
Centralized, Decentralized, & Distributed Networks(Baran
1962, Figure 1)
Positivist/Modern : Anti-Positivist/Postmodern :
Post-Positivist/UnmodernHempel/Carnap/Ayer :
Kuhn/Toulmin/Wittgenstein : Latour/Nersessian/Dewey
-
Ecosystem Alliances, Obligatory Passage Points,& Boundary
Objects
(Adapted from Star & Griesemer, 1989, p. 390)
Construct maps, models, specification equations
Instrument calibrations, standards laboratories, Wright maps
Data sources, kidmaps
-
Data
InstrumentTheory
-
chair
椅子la sillaкрісло
كرسي
Modern positivist extreme: One defined universal rules in
exclusive restriction
XX
X
-
chair
椅子la sillaкрісло
كرسي
Postmodern contextualist extreme: Local values rule in chaos of
incommensurate relativism
X
-
Metaphor, analogy, and mythology
• "Wittgenstein (e.g., 1967) claimed that many scholarly
confusions are induced by the "mythology in the forms of our
language", by metaphors and analogies that beguile their users, by
grammar projected onto reality, etcetera. Fields that involve
mathematics(both applied and pure) are traditionally rich ground
for the growth of language induced confusions, for symbolic
representations, while notable for their compactness, are
singularly able to mislead.” (Maraun, 1996, p. 603)
• Maraun, M. D. (1996). Meaning and mythology in the factor
analysis model. Multivariate Behavioral Research, 31(4),
603-616.
-
Metaphor, analogy, and mythology
• “Bertrand Russell speaks of cases 'where the premises of
sciences
turn out to be a set of pre-suppositions neither empirical nor
logically
necessary'; and in a remarkable passage, Karl R. Popper
confesses
very plainly to the impossibility of making a science out of
only strictly
verifiable and justifiable elements.” (Holton, 1988, p. 41; also
see
Polanyi, 1974, p. 323).
-
Identifiability in relation to semiotics
• "The strong hold that certain themes have on the mind of the
scientisthelps to explain his commitment to some point of view that
may in fact run exactly counter to all accepted doctrine and to the
clear evidence of the senses. Of this no one has spoken more
eloquently and memorably than Galileo when he commented on the fact
that to accept the idea of a moving earth one must overcome the
strong impression that one can 'see' that the sun is really
moving…”
• Holton, G. (1988). Thematic origins of scientific thought:
Kepler to Einstein(Revised ed.). Cambridge, Massachusetts: Harvard
University Press, p. 43.
-
Identifiability, semiotics, and models in relation to reality•
Cartwright noted, ‘fundamental equations do not govern objects
in
reality; they govern only objects in models” (p. 129).
• "Today we often consider things like electric and magnetic
fields to be real. This is not the case. The fields are purely
abstract mathematical constructs that allow us to predict things we
can actually perceive.“ (Rautio, 2005, p. 53)
• The point is whether the model-based governance of objects can
be made useful.
-
Economy of Thought:Bounded Rationality and Captivated
Imaginations
Georg Rasch and George Box:Models are not meant to be true, but
to be useful.
So what do we do when we repeatedly bump up against the limits
of our existing concepts, when the ways our imaginations are
captivated by the ruling metaphors of the day start to cause more
problems than they solve?
-
• Tocqueville’s (1805-1859) contrast of Europeans and
Americans.
• When asked, “Why do you do that?”• Europeans say, “Because
of…”
• Americans say, “In order to…”
• Both explanations are needed to establish basis for trust
indomains of economic intangibles.
How are allies convinced to trust a technology?
-
Why do you do things that way?
• “Because of…”• Lindquist (1953, p.35) "The objective [of an
educational test] is handed down by
those agents of society who are responsible for decisions
concerning educational objectives, and what the test constructor
must do is to attempt to incorporate that definition as clearly and
exactly as possible in the examination that he builds."
• “In order to…”• Wright (1977, p. 97) “When a person tries to
answer a test item the situation is
potentially complicated. Many forces influence the outcome--too
many to be named in a workable theory of the person's response. To
arrive at a workable position, we must invent a simple conception
of what we are willing to suppose happens, do our best to write
items and test persons so that their interaction is governed by
this conception and then impose its statistical consequences upon
the data to see if the invention can be made useful."
-
Thank you!
• William P. Fisher, Jr.• University of California, Berkeley
• [email protected]