Cognition, Information Processing, and Psychophysics

Basic Issues

Hans-Georg Geissler, Stephen W. Link, James T. Townsend, Hans-Georg Geissler, Stephen W. Link, James T. Townsend

• Publisher: Psychology Press
• Year: 2013
• ISBN:
  9781134994007
• Edition: 1
• Topic: Psychologie
• Subtopic: Geschichte & Theorie in der Psychologie

The plan for this volume emerged during the international Leipzig conference commemorating the centenary of the death of Gustav Fechner. The contributors suggested that while many features of modern psychological theory were anticipated by Fechner, many new theoretical approaches owe much more to him than often is realized. As such, they decided to honor Fechner by evaluating his own contribution to the founding of psychology and psychoanalysis, by deepening the foundations of psychological theories of consciousness, perception, and choice, and by using the analysis of time to create a new appreciation of constraints that bind mental processes together. Thus, this volume spans an extraordinary range of psychological topics, from hermeneutics to the time-quantum basis for mental processes, in a way that would both amaze and delight Fechner. Moreover, the international reach of his pioneering ideas can be seen from the current locations of the contributors. The span from Japan to the United States to Holland to Germany and to Israel provides a global measure of Fechner's scientific legacy.

I

FOUNDATIONS AND EPISTEMOLOGY

1 Fechner's Inner Psychophysics: Its Historical Fate and Present Status

Eckart Scheerer

Universität Oldenburg Federal Republic of Germany

ABSTRACT

Fechner's inner psychophysics, though at present completely forgotten, contained a program for a biophysically oriented psychology and occupied a central place in his thinking. Inner psychophysics rested on the following key assumptions:

1. The transformation between neural and mental events is non-linear.

2. There is a cascade of “inner thresholds” intervening between neural and mental events.

3. Neural processes operate on the oscillation principle and are to be studied by the methods of Fourier analysis.

4. The “psychophysical representation” of mental processes is highly parallel and distributed.

The reasons for the rejection of Fechner's inner psychophysics by his younger contemporaries are summarized and Fechner's program is discussed in the context of present-day cognitive science. Fechner must be credited with anticipating the holographic approach to brain function, and this may account for his insistence on linearity in the nervous system. However, there are also significant affinities between Fechner's thinking and modern parallel distributed processing approaches.

Today, Fechner's inner psychophysics is all but forgotten. Although every contemporary textbook of psychophysics pays its tribute to Fechner as the founder of the discipline, as far as I know, only two of them (Baird & Noma, 1978; Marks, 1978) mention his inner psychophysics; even more, it was already rejected in Fechner's lifetime by the vast majority of the scientific community. Yet to Fechner himself inner psychophysics was all-important; without it, he wrote (1882, p. 262), psychophysics was destined to be an “insignificant appendix to physiology.” If we want to celebrate Fechner and his work, we must not forget inner psychophysics. What did Fechner mean by “inner psychophysics”? What is the relation between his inner psychophysics and his psychology? Why was inner psychophysics rejected by Fechner's contemporaries? What significance could inner psychophysics have for present-day psychology? Those are the questions addressed in the present essay.

ASPECTS OF INNER PSYCHOPHYSICS

It is possible to discern several aspects of inner psychophysics which, though closely interwoven in Fechner's doctrine, may be taken apart and discussed separately. We begin our discussion with a comparison between inner and outer psychophysics and then proceed to a consideration of inner psychophysics in its own right.

Inner and outer psychophysics differ with respect to subject matter. Outer psychophysics considers the mental in its relation to the “physical outer world” and inner psychophysics in its relation to the “physical inner world, with which the mental is most intimately connected” (Fechner, 1860, Vol. I, p. 10). Or, to put it differently, on the way from stimulus to sensation there are two transformations. First, the stimulus (a process of physical motion) is transformed into “psychophysical motion,” Fechner's term for neural excitation.¹ Second, psychophysical motion is transformed into sensation, a mental phenomenon. Inner psychophysics deals with the second transformation, whereas outer psychophysics is concerned with the result of both transformations. From this standpoint, outer psychophysics is impossible without inner psychophysics, because psychophysical motion is an indispensable condition for mental processes to occur. On the other hand, those mental processes which are not immediately contingent on external stimulation are the potential subject matter of an autonomous inner psychophysics.

From the standpoint of methodology, the relation between inner and outer psychophysics takes a different form. Only outer psychophysics can be based on direct observation and experimentation (Fechner, 1860, Vol. I, p. 11); inner psychophysics draws inferences from outer psychophysics, “taking into account anatomical, physiological, and psychological facts” (Fechner, 1877, p. 3). Experimentation and quantification are impossible in inner psychophysics because psychophysical motion is not directly observable, and because mental phenomena are quantifiable only indirectly, via the physical phenomena to which they correspond (Fechner, 1851/1987a, p. 203).

Let us briefly pause to compare these constraints with the present situation in psychophysics. The first was removed by the advent of experimental brain physiology. But the second still serves to demarcate Fechner's psychophysics from some current mainstream approaches in psychophysics. “Direct” psychophysics, as inaugurated by S. S. Stevens, denies any fundamental difference between physical and mental phenomena concerning measurement and quantification. Other approaches, such as multidimensional scaling and functional measurement, rely on methods that are as indirect as Fechner's methods but do not require physical measurements. None of these approaches would have satisfied Fechner (Scheerer, 1989).

Given that sensory magnitude is a logarithmic function of stimulus magnitude, we must ask for the locus of the logarithmic transformation. As early as 1851, before he founded psychophysics on Weber's law, Fechner decided in favor of the brain-mind locus of the logarithmic transformation. This decision was never taken back by him. In 1863, Wundt argued that the logarithmic law resulted from the comparison between sensations, and in 1878 Georg Elias Müller brought together arguments in favor of a logarithmic transformation between stimulus and neural excitation. Against the “psychological” and the “physiological” interpretations of the logarithmic law Fechner kept defending his own “psychophysical” interpretation (thus named in 1882), which assigns the transformation to the realm of inner rather than outer psychophysics. The psychophysical interpretation also requires the assumption of inner thresholds intervening between psychophysical activity and conscious phenomena. Thus, stimulus (i.e., outer) thresholds and thresholds of consciousness (inner thresholds) are not the same, though the latter, too, must be inferred from measurements in outer psychophysics. At the end of his life (1887) Fechner concluded that the logarithmic law was superior to all other formulations in terms of empirical validity, but that in principle it could be replaced by the power law resulting from Plateau's (1873) “ratio hypothesis” of sensory magnitudes (Scheerer, 1987b). Thus, although Fechner himself did not spell out this conclusion, it seems that inner psychophysics restricts the quantitative relations between neural excitation and sensation to some nonlinear transformation.

Why is a linear transformation of neural excitation into sensation excluded? Fechner (1877) argued as follows. A distinction must be made between phenomena where the cause precedes the effect, and phenomena that occur simultaneously. In the first case, the effect is proportional to the cause; in the second case, the two phenomena typically are not proportional to each other, especially when they are aspects of a unitary event measured in two different dimensions such as space and time. Examples of such pairs of phenomena are the period (T) and the length (L) of a simple pendulum

or the period of a planet and its distance from the sun. In both cases we have a simultaneous, nonlinear dependency between spatial and temporal aspects of a unitary event. Relations between the physical and the mental are clearly of the second type. The physical and the mental occur simultaneously, and they are two aspects of a unitary event resulting in two different kinds of phenomena (see below). Consequently, according to Fechner they should not be proportional to each other. On the other hand, all events intervening between stimulation and neural excitation are of the first type, that is, they occur in succession and every effect immediately follows its cause; consequently, there is no known physical reason why they should not be proportional to each other. A more concise formulation of the same argument is that stimulation and neural excitation are governed by causal laws, whereas the relations of inner psychophysics are governed by functional laws (Heidelberger, 1988). In fact, psychophysics is defined as the science of functional, or dependency relations between body and mind (Fechner, 1860, Vol. I, p. 1), an expression that is not a mere way of speaking but that implies that psychophysical laws (in the sense of inner psychophysics) are of a special type not subject to ordinary causal laws.

In the last section of this essay, I indicate a strictly scientific reason, unfortunately not stated with sufficient clarity by Fechner himself, why he was forced to exclude nonlinearity from the physiological as well as from the mental domain and to reserve it for the relation between the two domains.

Though the mental and the physical are identical, in an ontological sense, the “radically different, inner or outer, point of observation” (Fechner, 1877, p. 67) produces an essential difference in their modes of appearance. The task of inner psychophysics is to relate these two modes of appearance, which differ from each other in at least three respects.

First, there is a quantitative difference, at least with respect to conscious mental phenomena. Consider first the case of the differential threshold, conceived as an inner threshold. If two brain states B₁ and B₂ cannot be discriminated from each other, then they are both mapped onto the same conscious state B*. Taken at face value, this argument implies that the mapping from psychophysical into conscious states is not one-to-one but many-to-one; although every psychophysical state has one conscious correlate, every conscious state has more than one psychophysical correlate. The fundamental indeterminacy introduced by this formulation was noticed in Fechner's last paper, in which he tried to remove it. On the other hand, the absolute threshold, if conceived as a threshold of consciousness, implies the existence of a continuum of unconscious (subthreshold) sensations, and thus the number of psychophysical states in any case surpasses the number of conscious mental states. Second, according to Fechner, psychophysical processes can and must be described in purely quantitative terms, whereas mental phenomena in addition possess qualitative attributes such as sensory modality or the difference between pleasure and pain. Third, these properties accrue to mental phenomena not as the result of some deficient mode of functioning, but in virtue of a positive principle: the mind's power of unifying the diversity that characterizes physical phenomena and converting them into a simplified appearance (Fechner, 1851/1987a, p. 205). This is one aspect of Fechner's Synechological Principle, to which we return in a moment. First, however, a brief comment should be made on the quantitative/qualitative side of the body/mind difference.

It seems that this aspect of inner psychophysics was considered by Fechner as at once the most puzzling and perhaps the most important. Most important, because if a reduction of quality to quantity were to succeed, it would put psychophysics on the same footing as “computational physics,” which had rendered “the quality of colors and sounds” amenable to quantitative determinations (Fechner, 1851/1987a, p. 203). Most puzzling, because Fechner was aware that within inner psychophysics the reduction of quality to quantity would remain for a long time purely hypothetical. For instance, in his esthetics he regarded it as a “fundamental need” to know the laws relating pleasure and pain to the bodily (“psychophysical”) processes underlying them, a task that would belong to inner psychophysics. Yet at the same time he had to admit that only more or less uncertain hypotheses could be formulated at this point. As a result, experimental esthetics would have to remain in the sphere of outer psychophysics (Fechner, 1876/1925, p. 48). Nevertheless, Fechner was prepared to put forward hypotheses linking quality to quantity. For instance, he considered pleasure to be the correlate of a psychophysical process approaching a stable state and unpleasantness to be the correlate of a movement away from stability (Fechner, 1873, p. 94)—a principle that was little heeded by experimental psychologists but was destined to gain great importance to Sigmund Freud, who also attempted to find out quantitative, “energetical” correlates of qualitative properties of mental life (Scheerer & Hildebrandt, 1988, p. 279).

Finally, inner psychophysics is concerned with elucidating the laws governing the most proximate correlate of mental activity, called by Fechner “psychophysical motion” or “psychophysical excitation.” In the present context, psychophysical motion may be equated with neural excitation. To Fechner, neural excitation was in part endogenous and consisted in mechanical oscillations. Its oscillatory nature is evident in phenomena such as sleeping and waking. There is a “main wave” of “psychophysical motion,” which is superimposed by subsidiary waves or “ripples.” All aspects of subjective experience are represented in “wave trains” of psychophysical motion, in parameters such as the amplitudes, periods, and phase relations of their spectral components. What is needed is a Fourier analysis of psychophysical motion. Fechner did not carry through this task, but the general idea is clear. Whether or not we notice a stimulus, for instance, is not a simple function of the intensity of the stimulus, but depends on the period and phase relations between the endogenous “main wave” and the exogenous “ripples,” and on factors such as the gradual or abrupt onset of the stimulus, its spatial distribution, et cetera (Fechner, 1851/1987a).

Why should organisms in general and neural excitation in particular display oscillatory motion? Initially, Fechner thought that oscillation was a simple consequence of the law of conservation of energy to which the organism is subjected. Under the law, energy cannot be added to a system without subtracting energy, perhaps in some other form, from it. Thus, an “antagonism” between two opposing forces is set up, and systems exposed to antagonistic forces tend to oscillate (Fechner, 1851/1987a). Later, Fechner (1873) proposed a more complicated hypothesis taking into account the difference between inorganic and organic matter. In inorganic molecules, the particles display caloric oscillation around stable equilibrium positions. In organic molecules, the particles are subject to inner forces and display a continuous irregular change of relative position. Nevertheless, even organic systems (provided they are closed or exposed to constant environmental influences) are characterized by a tendency toward stability, where “stability” is defined as a state in which “the particles return periodically, i.e., at equal intervals, to the same relative position and movement states” (Fechner, 1879/1904, p. 209). Under the influence of environmental forces, full stability is never reached but the system at least tends toward approximate stability by compensating for the effects of environmental forces. In virtue of these ideas, Fechner was an early but scarcely acknowledged pioneer of the Theory of Self-Organizing Systems (Heidelberger, 1990).

By means of the oscillation hypothesis, Fechner also got around a potential contradiction arising from his conception of mental activity. On the one hand, the mind is nothing else than the “inner appearance” of physical processes; on the other hand, the mind is presumed to be endowed with powers transcending the powers of its physical carrier. Fechner's solution was to give the synechological principle a psychophysical interpretation, which has a more molecular and a more molar side. First, mental activity is never carried by a single particle but always by an oscillatory system of particles, which at the very least requires two particles in motion. Second, mental activity does not reside in any single part of the psychophysical system but is carried by the entire activity of the psychophysical apparatus, which is spatially distributed over the entire brain. Or, to use terms not employed by Fechner himself, the mind is an emergent property of the entire psychophysical system and cannot be reduced to one of its (functional or spatial) components.

INNER PSYCHOPHYSICS AND PSYCHOLOGY

It is well known that Fechner had a low opinion of “pure” psychology. When Wilhelm Wundt informed him about his plan...