Phenomenological Commitments

The divergence on empirical, theoretical and methodological issues does not affect the core of the phenomenology of perception as it has emerged through the various philosophical and scientific theories in which it is embedded. This core consists of the following minimal set of commitments.

Perception is an independent cognitive function. If considered from the standpoint of the things in the world to which it refers, it can be studied as a mode of cognition under the respect of the accuracy with which it makes them accessible to subjects. However, perception conveys this information by means of a form that has the same function as grammar has for language. It corresponds to factors and rules that underlie the features, the structures and the connection of appearances, which do not depend on other cognitive functions or the material properties of things. Factors and rules can be derived from appearances considered in themselves, that is, through descriptions that do not mention philosophical constructs and scientific findings on their extra­perceptual causes that are not required by a complete account of what appear­ances afford within the limits in which they present themselves. For instance, Hering contrasts his view with that of von Helmholtz by treating colors as the independent properties of things in naive experience, but he introduces the construct of “visual things” to study the building blocks of the perceivable world regardless of commonsensical or scientific beliefs in their existence. Likewise, Rubin claims that the devices and the visual displays of the research should have the same complexity of things in ordinary experience, while at the same time he deals with the features of figure, ground and contour as purely phenomenal data. Rubin emphasizes that were these features different, there could be no successful understanding of the environmental world, because the distinction between enclosed and enclosing regions, the delimitation of forms and their independence of location, the continuity of surfaces behind one another would not occur. Thus leaving out of consideration the things of the external world as such and the knowledge function of perception means de­parting from the naive realism of the common-sense associated with ordinary experience, according to which any question about perception is exhausted by the material properties of objects or by what one believes is acquired from objects by learning. Michotte gives a clear example of the extent to which the commitment to the independence of perception means studying what the world presents to subjects just by limiting the research to the form of percep­tion. Contrary to the prejudice that could stem from philosophical or psycho­logical theories, ordinary experience shows facts whose proper understanding involves the perception of causality. Then causality perception is investigated by designing systems of stimulation that enable the separation of phenomeno­logical structures from beliefs and expectations based on commonsensical or scientific knowledge.

It is consistent to maintain that perception has to be explained at face value, namely as it occurs, without being laden by theories in the experience of the world, and that the common-sense of naive experience must not be admitted in descriptions. In fact, the aim is to capture perception as it is instanced in the direct experience of subjects in order to explain it independently of any biased theory, be it the naive causal ascription of perception to the material prop­erties of objects or the sophisticated knowledge of philosophy and science. This aim is clear in Brentano and from Husserl’s concept of the “natural world” it extends to Kanizsa’s criticism of the naive realism of the ordinary life com­mon sense (on common-sense cf. Smith, 1995b). Accordingly, the phenome­nology of perception construes external objects as perceivable things, qualities and quantitative properties that are considered the bona fide referents dis­played in manifold appearances. In this sense, Husserl ([1952] 1989: 91-92) defines things to be the intersubjective rule of their appearances. Kohler (1938: 70) holds that “the question of phenomenology” is “what properties the thing actually has.” For this reason, phenomenology does not accept the reduction of appearances to aggregates of sense data or to the subjective seeming of the external world as it is defined by philosophical and scientific theories. From the experimental viewpoint, Katz’s research on color and tactual modes of appearances of things is an outstanding instance of this epistemo­logical tenet.

Consequently, another phenomenological commitment is that perception has a content resulting from the abstract distinction between outer objects and perceivable things or the former as appearing referents. The content is the specification of what, both in the perceivable world and in the displays used in the experimental conditions, subjects can ascertain and point out as indepen­dent of and resilient to their cognitive integration, expectations and change of attitudes. Appearances are considered as taking place in the field outside of subjects and “including” the things and properties as what they point to, or rather implying what of these things and properties falls into perception. The elements and connections that are specified in the content count as real just because they are forced upon the subjects and have perceivable effects on one another. The elements that cannot but appear to belong to one another in Meinong’s analysis, the phenomenal characteristics in Metzger’s theory of what is encountered, and the constraints on the changes of independent or dependent parts are all examples of this sense of perceptual reality. For this reason, real appearances also display a “natural” trait. This natural trait is not epiphenomenal or parasitic on what it is like for subjects to experience how something looks. It is associated with the fact that their independent and re­silient elements build a collection in which they occur in connection with one another. Grossmann (1977: 6) captures this fact through the concept of struc­ture. Let the perceivable things A and B respectively be green/round and red/ square. If they were described as sets of properties, there would be more sets than things, like for instance the set {round, square}. Yet this would provide false or at least poor descriptions of A and B that miss the inherent connection they show in experience. This is the sense of Rubin’s observation that trying to see the contour connected to the ground is not only difficult but also “strange” and “unnatural” in ordinary experience. Likewise Wertheimer claims that a grouping is natural for a perceptual scene against all possible aggregates into which it could be decomposed. The subjects’ latitude in eluding the percep­tual properties and connections forced upon them is indeed constrained by the admitted possibilities given one or more perceptual structures. This obser­vation is decisive for reports on color and shadow settings delivered by Her- ing, or subjects protocols in the experiments of Wertheimer and Goldmeier. Indeed, however great an effort a subject can make to change her cognitive attitude or refer to the knowledge of the system of stimulation, she will not be able to see an instance of the launch effect as the movement of two indepen­dent objects.

Another phenomenological commitment is that the observable elements of the phenomenal content are captured through the primitives and the experi­mental variables of theory in order to discover the structures of appearances through analysis and functional connections. Since Brentano, the analysis has consisted of making the inherent characteristics of appearances explicit and treating them as abstractly separable phenomena. Analysis decomposes perception into its constituents to set out the elements that can bear an in­dependent variation along with the rules of their connection. The analysis is equivalent to an explication of what is already contained in perception such that anyone can decide whether and how the specified elements and con­nection match the perceptual meaning of which one has a direct experience. The phenomenological analysis is different from the concept of explication set forth by Carnap (1950: 3, 6, 7), which is the replacement of the vague and inexact concepts of everyday language (e.g. fish and warmth) or of an earlier stage of science by the concepts defined by the explicit rules of a well-constructed system of either mathematical or empirical concepts (e.g. piscis and tempera­ture). This kind of analysis is not a clarification of meaning by making explicit the characteristics included in the replaced concept. It is the reconstruction of the “pre-systematic” concept; hence, although it should be similar to some extent to the replaced concept, the reconstructed concept might diverge from the intuitive meaning of the replaced one. Instead, Husserl’s bottom-up speci­fication of meta-theoretical concepts of phenomenology aims at recovering the inherent properties of perception that fix its meaning as it occurs in ordi­nary experience, because they grasp the invariant features and structures gov­erning its internal coherence and consistency. The analysis of philosophical psychology that Hering demands for the experimental research has the same aim. In this sense, the neutral character that Stumpf ascribes to phenomenol­ogy is an ideal passage from analysis to an experimental practice whose vari­ables are designed by means of selecting pieces of the perceptual scene and crafting the conditions to derive their underlying yet still observable rules. The procedures that Wertheimer, Kanizsa and Michotte have followed to study the grouping factors, the amodal completion and the structure of causality are examples of how the perceptual rules are derived from the interdependence of directly observable experimental variables. At the borders between theo­retical and experimental abstraction from the content of perception, there is the construction of the series through which Hering observes the appearances “from the outside” and Goldmeier ranks the variables of phenomenal simi­larity by testing the observable effects of its change in visual figures. Hering wants the theory to represent the essential properties that one can observe to define the appearances. Goldmeier needs a paradigm to control the phenomenal magnitude and direction of similarity change. Musatti and Metzger show how the experimental research on spatial perception can be carried out through transformations induced on appearances. Thus the con­tent of perception is the basis of abstraction of repeatable phenomena. Koffka (1924: 155) emphasizes that any phenomenal reaction of the organism to the environment, like perceptual appearances, can be studied by the substitu­tion of adequate theoretical constructs or transformations induced in experi­mental conditions. However, “the original reactions are to be studied just as they are and not merely under the aspect of what they will become when the analysis is applied to them. Only in this way shall we able to find their proper laws.” The theoretical primitives, the selection of the experimental variables, the explication of meaning and the controlled transformations should be the most suitable to abstract the structure that explains the course of appearances, rather than being derived from beliefs, theories and evidence that refer to non­perceptual entities.

This demand is consistent with the phenomenological commitment to dis­cover the de re properties of perception. A de re property can be expressed in modal terms as follows:

where the quantified variable does not occur within the scope of the modal operator of necessity (□); hence, (1) expresses the necessity of referential sen­tences, whose truth is determined by the relation between terms (see Fine, 1978: 1, for this formalization). An instance of a phenomenological proposition on perception whose truth is obtained through the analysis of the observable referents denoted by the terms is:

where F and G stand for features like color and surface, boundary and figure, good continuation and launch effect, pitch proximity and melodic figure or tonal quality and fusion. This means that a color appearance necessarily brings about the need for a surface and the demand for the latter to be filled by a color; or that a boundary-contour necessarily delimits something along one di­rection and a figure necessarily demands to be enclosed. The necessity regards the observable elements of the connection so that one could still conceive of a world in which a color does not need a surface or a figure is not enclosed by a boundary, but if this connection holds in perception it necessarily im­plies that the colors in the perceivable world must spread on surfaces and the boundaries must delimit asymmetrically an enclosed figure. Instead a de dicto property is expressed by the formula:

in which there is no free variable within the scope of □; hence, each part of (2) is also a sentence so that (2) expresses the necessity of sentences whose truth is determined by the connection with one another in a system of propositions. The distinction between a de re and a de dicto interpretation of a property is like that between saying “x is necessarily F” and “that ‘x is F’ is a necessary proposition.” An example of a de dicto interpretation of a property of percep­tion is:

in which ◊ is the modal operator for the possibility, x is a chord, F stands for a definite interval and G for fusion. In this interpretation, it is impossible that all the chords with tones in the F interval do not give rise to fusion, but this does not depend on the perceptual nature of fusion, rather on the function that in­tervals of the kind F have in a tonal system or a musical culture. The necessity does not hold between the perceptual properties of the tones in the chord and the perceivable fusion, but it depends on the system of rules to combine tones. A chord that, notwithstanding the appropriate interval, did not bring about fusion would only be a chord that does not make a functional difference in the tonal system to which it belongs. Instead, fusion is a de re property. This does not mean that there are no functional differences due to musical cultures, rath­er that these are a specification of the perceptual properties of tones given the constraints of a system that may select and exploit some of them (cf. Stumpf, 1911; Tenney, 1988).

The features, the structures and the functional connections that figure in phenomenological descriptions are necessary de re properties of the observ­able content of perception, which are derived through the analysis or the ex­perimental research. Thus the emphasis on the phenomenological description does not imply a sort of opposition to the explanation, as if phenomenology provided an account that claims to differ from standard logical and scientific accounts. The relations of dependence analyzed by Hering and Stumpf and conceptualized by Husserl, the characteristics of the modes of appearance discovered by Katz, the forms of connection of Brentano and Rubin as well as the grouping factors of Wertheimer are de re properties which capture the compelling characters of appearances and correspond to perceptual rules.

Far from derogating from explanation, the description has a twofold function. It represents the abstracted properties of appearances and accounts for them in terms of interdependence of observable factors and self-consistent rules. Phe­nomenological descriptions denote instances of the perceptual understanding of the world and at the same time fragments of the grammar of perception.

Therefore the phenomenology of perception is not a theory of the first- person experience that needs the science of perception to recast his methods of collecting and analysing data in the third person. Nor is description a re­construction of naive theories or platitudes in the sense of folk psychology or the sophisticated report of trained subjects about their mental states, be they qualia or attitudes, to experience a task in controlled conditions (cf. Stich, 1983; Stich and Nichols, 2003; Varela, 1996; Lutz, 2002). Duncker (1932/1933) main­tains that the theory of perception is not interested in subjective experiences, rather in the forms ruling the course of appearances on which the order and implication of the propositions of the theory depend. The theory of color per­ception deals with the properties that order the phenomena within the color system. The subjective experiences of color blindness are important insofar as they permit the study of the differences between normal and pathological color systems and the discovery of whether the passage from one to the other preserves the order and implications of propositions on color phenomena. The difference of experiences is equivalent to a transformation through systems that is relevant to see whether or not the equivalence of propositions is pre­served for both. The difference might leave the order and implications of the propositions unchanged, like when Hering’s color series is observed from one direction or the other or when the perceptual constancy of things is restored after wearing distorting spectacles. In similar cases, first-person experiences are like models in which the interpretation of the theory is true. It is possible that differences between subjects change the properties of propositions from one system to another. Yet such changes are anchored to the regular course of appearances so that it is possible to test whether subjects use the same words, for instance “greater than” and “right neighbor to,” to mean the same ordering of appearances.

As Husserl (1900/1901) and Koffka (1921) remarked, each theory builds its own phenomena, because it selects its primitives, constants and variables in connection to its own domain of observables. Consequently, perceptual phe­nomena are not the subjective contents of first-person experience, rather the abstract construct of the theory dealing with the features, the structures and the connections of appearances through the analysis and the experimental testing given a number of observations repeatable at will. A phenomenon is the observable array of structural properties, rather than the experience of each subject at a given time, which accounts for the referents of possible appearances. The abstract character of phenomena allows one to treat their structural properties as the lawful content of the rules of the perceivable world. In ideal connection with Husserl’s and Kohler’s claim that phenom­enology deals in reality with things, Bozzi (1978) defines a perceivable thing as the “logical sum of the possible configurations given the arrangement of its observable elements.” This definition holds true for the experimental evidence on the structures of movement, phenomenal causality and figure – ground as well as for the analysis of continua or of the coordinate systems of space, as it was carried out respectively by Brentano or Meinong and Husserl. In these phenomenological theories, phenomena are theoretical units equivalent to what is perceivable in possible, namely repeatable, experiences, with which what is perceived from a first-person standpoint in single circumstances has to be consistent. For this reason, Hering admits of pure colors as the ends of the series of appearances, while Brentano and Meinong conceive of an indefinite extension of tones or colors. Under this respect, Husserl’s eidetic variation and Michotte’s concomitant variation could be considered as the theoretical and the experimental sides of the same phenomenological method (Thines, 1991: 17; cf. Albertazzi, 2013).

Phenomenology has a no less abstract theoretical structure than do other theories and sciences. It is not a method for supplying psychophysics or theo­retical psychology with qualitative data interpreted as how a stimulus seems to subjects under a respect (cf. Horst, 2005). Nor is it merely a philosophy that faces the question of how its concepts could be integrated with the standard account of science (cf. Gallagher, 2003; Gallagher and Sorensen, 2006; Over- gaard, 2004, for a discussion of the modes of integration). To be sure, in the literature on such issues it is argued that qualitative data are not theoretical posits, namely unobservable constructs that are introduced only to account for data of a different kind, but intersubjectively reliable data. Yet, this literature misses the evidence of experimental phenomenology and the consistency of phenomenological analysis with science that is due to the explanatory func­tion of description. Furthermore, what this literature does not acknowledge is that phenomenology deals with the qualitative as well as the quantitative dimensions of perception that are shown to have an order and that can be treated as magnitudes, thus providing an autonomous basis for formal models and measurement.

A commitment of the phenomenological theory of perception is that the perceivable world consists of forms of order at scales that span the connec­tion or discontinuity within and among things, the distribution of things and qualities in phenomenal space and time as well as the forms of the latter, the relations that hold the elements together in the modes of appearances of color, touch and movement. The analysis of the perceptual concrete manifolds and the theory of parts and wholes, the so-called mereology, is the common thread from which the phenomenological research into the order of appear­ances derives. The concrete manifolds correspond to collective properties and are denoted by names like: pile, storm, flock, herd, bunch, heap and so forth. If the partition of such manifolds is taken into account to classify them, there emerge important differences in the mode in which their constituents are held together. If the manifold is indifferent to relations among constituents, then it is an aggregate. If the aggregate is made of constituents of different kinds, then it is a sum. In such cases the constituents are parts on a par with the whole they build. In fact there are many kinds of aggregates and nothing excludes that the parts are somehow tied together or that the manifold is specified by a property that is not possessed as such by all the parts. Yet the rule of composition of the aggregate denotes a combination extrinsic to the parts as well as that the property that specifies the manifold is not subject to a restriction on the na­ture of parts, so that it may bring about a scattered or ill-formed whole at least on intuitive grounds (cf. sum and “fusion” of classical extensional mereology; cf. Varzi, 1996). There are instead cases in which the manifold is not perceived as a whole of disparate parts or a sum. For instance, in a flock of geese each goose is part of the flock and the flock is the plurality of the parts (von Eh- renfels, 1890; Husserl, 1891; Meinong, 1891). The parts neither build a set or a sequence of similar elements nor hold an external relation to one another, rather their collection displays a characteristic unifying quality that makes the parts interdependent and is “grasped in a single glance” (Husserl, [1891] 2010: 216). Such a quality may be quantitative, as in the perceptual categorization of the quantity of a finite group of elements (Husserl, 1891; Wertheimer, 1912b), or qualitative like the melody unifying tones that remains the same through the transposition from C major to F sharp. The unifying quality belongs to the manifold because it is founded on the nature of its parts. It rests on the fact that an avenue of trees is a row of elements without qualitative difference, so that their collection is a unity that shows no internal discontinuity, as well as that the tones of a transposed melody are such to preserve their intervals.

On this basis, the nature of concrete manifolds can also be specified un­der the respect of the observable changes they can bear. If one can either shift from the parts to the whole or arbitrarily add and remove parts in a con­crete manifold of constituents of the same or different kinds without causing changes, the manifold is an aggregate in the sense of a sum. If the aggregate or the sum has a rule of composition that assigns a position to the constituents but their relative positions are independently variable, the manifold is an aggregate in the sense of an “additive grouping” (Kohler, 1920; Rausch, 1937, studies the types of additive aggregates). Albeit in different ways, the parts of the aggregates and the sums are really separable from one another and each of them is detachable from the whole, because the form of their collection is indifferent to their nature; it persists in spite of thoroughly arbitrary varia­tion (Husserl, 1900/1901). If the manifold constituents cannot be arbitrarily separated or detached and the possibilities of variation are constrained, it will correspond to a narrow sense of whole whose parts are not disconnected or intrinsically unrelated. This holds true for the whole with the unifying qual­ity as part. If the types of “separability” and variations are taken into account, the kinds of the parts and wholes can be studied. If the variation of all the contexts in which a part occurs does not change that part, it can be separated and stand as an independent part. Then the part is a piece because its appear­ance does not need any other part; hence, it need not be included in a whole (Husserl, 1900/1901). If the variation of all the contexts in which a part occurs brings about a change of that part, then the part cannot really be separated and is a dependent part in the sense that it cannot stand alongside another co-occurring part as if they were two adjacent pieces. This part is called “mo­ment” to emphasize that its appearance needs other parts or a whole of which it is one of the parts (Husserl, 1900/1901; Kohler, 1920: 32). Pieces may build a whole in a narrow sense that is yet different from an aggregate, when they are tied together by just fitting one another, as in a jigsaw puzzle, or by another independent part (cf. Wertheimer, 1922, on this “additive” unification). Instead, moments always build a whole through unilateral or mutual interpenetration (durchdringen), for example that holding for that color and that extended sur­face of this table. Stumpf (1873: 113) argues that such parts are inseparable, or at least only abstractly separable, and that they form necessarily a whole because the smooth decrease and disappearing of the one causes the smooth decrease and disappearing of the other without any qualitative change of both. Were color and extension a sum, the disappearance of the former could only cause the disappearance of the latter. Then color is a dependent part or a moment of the surface. Likewise, the characteristic unifying quality of an avenue of trees is as dependent a moment on the trees that it unifies as is the melody on the tones, because they both cannot appear without the trees or the tones, while they appear if the latter have suitable properties (cf. von Ehrenfels, 1890; Meinong, 1899; Gelb, 1911; Wertheimer, 1922; Kohler, 1938; Stumpf, 1939; Rausch, 1964).

The relations of dependence and independence emerging through varia­tion specify the nature of the parts but also their connection in the whole, hence the inherent ordering and nature of the wholes (to account for whole as such, the mereological concepts need to be integrated with topological con­cepts, cf. Smith, 1996; Varzi, 1996). Brentano (1979, 1982) distinguishes between separable and distinctional or abstractly separable parts. Separable parts are independent of the whole, namely the whole is founded on the parts that make it up. Distinctional parts are dependent on the whole, namely they are founded on the whole in which they are distinguished. Such parts need to be connected in a whole, like a surface to a three-dimensional solid. Indeed, con- tinua are defined according to the separability of parts, namely the ordered partition allowed by the primitive “being part of” and by topological concepts like coincidence, boundary, connection. The relation of “belonging to” is not the equivalent set-theoretical concept. Parts and wholes are not different kinds of entities, like point-elements and sets, rather they are entities on a par among which relations of dependence hold (cf. Smith, 1992/1993; Baumgart­ner and Simons, 1994; Libardi, 1994). Boundaries necessarily depend on higher­dimensional continua, while if a boundary is itself a continuum it necessarily has continuous parts with the same dimension, that is to say it cannot be par­titioned in lower-dimensional boundary points. Things are closed continuous objects whose boundaries, surfaces and inner parts belong to one another at various dimensions. Husserl (1900/1901) employs the concept of “foundation” to define a more narrow sense of whole than those captured by the irreflexive property of “being a part of” or by the narrow sense of a plurality with a unifying quality. It captures the intuitive notion of things that are perceived as absolute­ly independent wholes, that is, self-enclosed wholes whose parts are unified by one another within their common boundaries that set a discontinuity with their surroundings. He denotes this kind of whole as the “pregnant concept of whole” (1900/1901: 282). Every moment needs to be combined with other parts, so that Husserl admits always the combination (Verknupjung) of at least three things, of which the first two are disjoint but belong to the third. Yet it is not always true that any combination of parts amounts to a well-formed thing. The pregnant concept of whole corresponds to things that do not appear as the combination of two or more really separated objects, since every part is found­ed on the others or, equivalently, each part is directly or indirectly connected with every other through foundation. If one part of the whole is removed, the other parts continue to depend on one another. Thus foundation is the core of the unity of things: a part of a pregnant whole needs to be supplemented by the others, and this is possible because of a single foundation that connects all the parts together. A thing is a whole that occurs if the supplementation of the parts is satisfied, if there are dependent parts that cannot exist in less inclusive unities than those in which they are connected. The requirement of the single foundation permits the avoidance of the indefinite regression that arises if the connection is due to a “formal part” that is included in the whole alongside the connected parts, as in Twardowski (1894), who classifies the nature, rank, degree and order of parts and relations in a whole.

This kind of analysis sets a common ground for the phenomenology of per­ception. The concept of “natural part” (Wertheimer, 1922) and the definition of the relations allowing a part to appear self-sufficient, incomplete, astray, unbefitting when isolated from the whole (Rausch, 1966) stem from it. Kohler (1925) remarks that decomposing a whole in the opposite direction to the or­der of parts given by their dependence relations faces a resistance in the in­verse direction by the whole. This remark is justified in the light of Husserl’s concept of foundation. If the connection of parts through a single foundation makes up a pregnant whole, there are parts that are immediately connected to one another, because they will have at least a common boundary, and parts that are connected only through the combination with intermediate parts. For instance, brightness cannot be immediately connected to a surface, because it requires that the color to which it belongs is spread on the surface. Conse­quently, if each part has a neighborhood of other parts, each part will not be at the same distance from every other and the partition of such a whole will follow an intrinsically ordered progression. On the contrary, a whole made of pieces does not display a necessary order in the sense that there will always be diverse potential partitions that make the same part explicit, since there is no ordering relation that gives a piece its place in the whole. The well-known research of Benary (1924) on the brightness contrast shows the perceptual ef­fects of being a part. Consider two identical grey triangles of which one lies on a larger black triangle, so that its hypotenuse is coincident with a border of the large triangle, while the other lies on a white ground with its catheti coincident with the borders of the two arms at right angles of a larger black cross. The hypotenuse of both triangles is adjacent to the white ground, but the triangle that lies within the cross arms has a larger neighboring black region since the cross might contain the larger black triangle. Nevertheless, the grey triangle adjacent to the cross appears dimmer than the other. It undergoes less the con­trast of the more extended black region, because it does not belong to it, than the grey triangle inside the large black one, which instead appears brighter although surrounded by a less extended black region.

The phenomenological theories assume that these kinds of part – whole prop­erties and connections build the furniture ofthe world as it is experienced through perception.^[1] They provide the phenomenal grounds for models, formalization and measurement. Kohler (1938: 43) criticizes Kant for the conviction that the “prerequisites” of science stem from the categories of mind, once it is assumed that “in the ‘material’ there is no basic principle of order.” Instead, perceptual orders allow one to construct the algebraic structures and abstract manifolds to map features and relations of appearances.^[2] The principles of order and functional connections captured by analysis and experimental research can be subject to formalization. The formalization may consist in the logical defini­tion of the qualitative unifying moment (Grelling and Oppenheim, 1938, 1939; Rescher and Oppenheim, 1955; cf. Simons, 1988). It may be the theory of the “ma­terial axioms” about the part – whole relations of phenomena, as in the semi­formalization of Husserl’s theory (1900/1901) (cf. Null, 1984; Fine, 1995; Casari, 2007, for a topological reformulation; Simons, 1982, for a modal interpretation). The formalization can also be meant to build a pure theory of part-hood, name­ly of the concept of part and of the partial or the strict partial orders based on it, of connection and so forth.^[3] Although this kind of formal theory is free from instantiation in particular cognitive or ontological domains, the latter might be considered models of the theory.

The perceptual principles of order can also provide the intuitive basis of measurement. The evidence of experimental phenomenology has shown that there is neither opposition between qualitative and quantitative dimensions in perception nor inconsistency between phenomenology and mathemati­cal or geometrical models. Brown, Michotte, Bozzi and Rubin worked on the quantitative appearances in common experience. Kohler remarked that the qualitative – quantitative divide, if sound, arises in the perceptual experience. Brentano and Stumpf admitted of phenomenal magnitudes, and on this basis they proposed to reform psychophysics and contributed experimentally to it. In addition, Meinong (1896) showed that the concept of measurement itself is consistent with phenomenology because of the ordering properties of phe­nomena. He contends that a quantity in a general sense is obtained each time qualities of the same kind are put in a linear order. There is no ontological gap between qualities and quantities, because the latter are neither entities of a particular kind nor posits of the conventional definitions of measurement. The operations of measurement are carried out on things from which an abstractly separable feature is extracted and treated quantitatively so that all features of the same kind are compared under a selected respect with a standard unit. Meinong (1896: 219-220) defines magnitudes as any x that in a multiplicity of x allows interpolating values between x and non-x, which fall in the same di­rection as non-x. Thus, for instance, any tone affords a magnitude under the respect of loudness whenever less loud tones are interpolated between any tone and silence. Since magnitudes are linearly ordered in the same direction towards the null value, their quantitative relations are equality and diversity. Equality means occupying the same place in the ordering series. Given the same direction, diversity means occupying a place that divides the series into two sections, so that each of two diversely loud tones either precedes or suc­ceeds the other one. Therefore, between two dissimilar tones there is a distance corresponding to a proper part of the ordering series that separates them. Dis­tance is a magnitude as well, because it is possible to interpolate shorter dis­tances between each distance and its reduction to zero. However, the distance between two points in an ordered series cannot be divided into smaller dis­tances. It is a relation that takes the points as terms; hence, it cannot be divided into smaller entities that are still relations. Instead, the segment conjoining the points is a stretch, which may consist of spatial or temporal units, that has other smaller stretches as components into which it is divisible. Each smaller stretch of a segment is still a segment that is infinitely divisible. In order to de­termine the distance, it must be coordinated with a spatial or temporal stretch whose subdivided parts of the same kind can be compared (1896: 232, 278). For instance, if the ends of a pair of compasses coincide with the points A, B, C on a line, the distance between A B and B C is equivalent to the diversity of posi­tion of B and C, while the difference between A B and B C is the segment B C (1896: 300, 303f.). This distinction is essential for measurement that, according to Meinong, is an indirect method to compare magnitudes (1896: 271-273). In general, comparing magnitudes is not limited to assessing equality or diver­sity, rather it can determine which value has each feature occupying a place in the linear ordering as a member of the disjunction “equal to, greater or less than” (1896: 236, 245-246). For this reason, one has to compare the subdivided parts of the stretch, which is coordinated with the distance between diverse features, in order to determine a difference. As the example of the line shows, the difference requires the partition and divisibility into comparable smaller units that are to be coordinated with distances. Measurement is an indirect method of comparison of parts through the assignment of numbers, which are infinitely divisible abstract magnitudes, to distances. By using abstract magni­tudes, measurement also allows the substitution of the magnitudes of a kind X to determine distances of a kind Y, for example when temperature is gauged by the length of the mercury expanding or contracting in a thermometer.

Meinong’s theory of measurement has advantages and shortcomings (cf. Tegtmeier, 1981, 1996). At any rate, it shows that qualitative and quantita­tive phenomena can be treated as magnitudes and measured, once the proper coordination is established between the method of measurement and phe­nomenological properties. Unlike conventionalist theories of measurement, for which quantities are the empirical relations defined by measurement while equality and diversity are logical relations, Meinong holds that the entities quantitatively treated and the linear order are given independently of mea­surement operations. His theory is surely consistent with the “representational theorem” of the model-theoretic account (Suppes and Zinnes, 1963), since it ascribes to measurement the function of replacing continuous entities with the discrete and precise representation by means of numbers, whose relations are assigned to the relations among the former in a coordinated manner. Yet these relations are not induced by the numerical structure, rather they are ap­prehended in a generalized and reliable way because they emerge by putting the features of things in the linear ordering (cf. also Witte, 1960, and on the measurement of absolute perceptual properties Witte, 1961).

Therefore, nothing in principle excludes that features and dimensions of phenomena, which have been accounted for by phenomenological descrip­tions, may be coordinated with a mathematical representation for measure­ment. The project of the intrinsic geometry of phenomena, which was outlined by I. Kohler and Rubin, and the phenomenological reform of psychophysics by Brentano and Stumpf are outstanding examples of what a “metric” with phe­nomenal standard units and magnitudes could be like. Yet regardless of Mei- nong’s tenet that quantitative relations are fiat or bona fide empirical results, one may raise the objection that the definition of the standard unit, whose concatenation is represented by the numerical relations, is conventional. How­ever, Musatti (1959) argues that the functional relation that defines the stan­dard unit and the measurement operation may well be conventional but not arbitrary. For instance, time is defined as the quantity that is proportional to the equal spatial stretches covered by the motion of natural or artificial clocks. Equal times will correspond to equal spaces covered by the uniform motions of the apparent displacement of the sun or of the hands of a clock. Still, the fact that the motion is uniform cannot result from measurement, because this would imply the very functional relation that establishes time as proportional quantity. This holds true also for the body chosen as measurement instrument, because it remains identical to itself through the concatenation of motions that bring it on the objects, which at the same time cannot but be the standard unit of measurement. In fact, no movement nor any body can be reasonably chosen as instances of uniform motion or rigid bodies. Musatti argues that there must be “phenomenal indicators” of the mobiles and bodies on which the functional relation can be founded. There are movements without iner­tial forces among those that a thing displays, which can then be assumed as uniform on the basis of observation. Likewise, there are indicators that per­mit one to select a more and more restricted class of bodies whose number of transportations to cover the whole extension of another body can be assumed to be proportional to its length. For example, in the perceivable world there are quantitative characterizations of the extension of things through which they appear equal to, or either greater or less than, one another. Moreover, the prin­ciple of free mobility is based on the observation that it is not true that every thing is deformable; hence, for some perceptual things the expectation arises that if put side by side their ends either tally with or extend beyond each other. Consequently, it is possible to observe relations of order that are roughly pre­served when one passes to the construction and definition of the instruments and units of measurement. The phenomenal indicators account for the fact that when a measurement result is consistent with the expectation that two apparently equally extended things tally with each other, this does not come out as a “surprise” (Kohler, 1938: 148). The functional relation defining a con­ventional standard unit is founded on the equivalence that holds between per­ceptual ordering and measurement at least up to some kinds of displacements, rotations and reversal of the unit chosen as standard for the comparison with ever increasing arbitrary nature. Far from allowing the suggestion of a wide gulf between phenomenology and the standard of science, this equivalence permits the autonomous logic of phenomenology to be mathematically rep­resented and measured. In general, the observation of the difference between physical length, as it is defined by the conventional measurement operations, and perceived extension in particular conditions is meaningful only if one can already perceptually compare the extension of distinct things and then apply to them the measurement instruments. Otherwise, the observation of the dif­ference between perceptual and physical values would make no sense (Kohler, 1938: 148, 151, 157, talks of the “reasonableness” of the measurement operations, even restricted to the detection of coincidence). This coordination provides the basis for the epistemological argument that the meaning of the units and the symbols of measurements carried out on the physical properties of the stimulation stand for the operations that are to be made on the stimulus ar­ray to allow for the independent structures of appearances to be established, pass to or be removed from one another. In this connection, phenomenologi­cal descriptions capture structures and rules that count as de re intersubjective truths of perception. Phenomenology as a theory embedded in philosophical analysis or as experimental practice is not only compatible with, but is fully consistent with, science in the third person. It requires only that features and dimensions of phenomena not be replaced from the outset by geometrical and mechanical properties resulting from measurement operations. However, it admits of a coordination that does have to be not arbitrary, but this is as sound a demand as the connection between mass and weight (Meinong, 1896: 262).

This consideration leads to a final assessment of the construct of “percep­tual reality” (Epstein and Hatfield, 1994). The difference between the results of measurement operations and the repeatable features of phenomena is as ob­servable as the inconsistency that emerges, for instance, between the appear­ance of causal movement and that of two independent unrelated movements. Consider, for instance, the perceptually unequal extension of the two lines in the Muller-Lyer figure and their equal length measured with a ruler. At the level of such a basic measurement operation, unequal and equal length can both be captured by two distinct Cartesian propositions as two incompatible observables (Bozzi, 1976: 159). The science of the physical constituents of the world discards the phenomenological description, which turns out also to be inconsistent with successive physical measurement operations, and assumes the maximal invariance of the physical measure across perception (Kohler, 1938: 149-150; Musatti, 1959: 368). Perceived inequality is no longer considered as an abstracted piece of the perceivable world that may figure in the proposi­tions of the science that aims at reconstructing the grammar of perception. Rather, this observable is discarded and the perceptual state of affairs is spoiled by the properties at variance with the physical invariance that becomes the bearer of the material properties, which the physics of macroscopic objects ascribes to the common-sense world. These properties can figure in the causal theories of perception, but can also be used to design pieces of the stimulation array simply because of their physical invariance.

In this connection, it is evident that the phenomenology of perception is not a phenomenalist theory, that is to say a theory that admits only the phe­nomenal content of single experiences of numerically distinct subjects, on whose grounds it is possible to posit the material things of common-sense and the physical objects of natural science considered as practical or logical constructs. The phenomenal world, or in Metzger’s terms the world of what is immediately given and encountered, is the world appearing according to the form of perception, which has to be accounted for by a neutral description with regard to the ontological commitment of physical sciences. As far as the correlation is concerned between the perceivable world and the world as it is studied by the natural sciences, phenomenology may be consistent with a kind of critical realism. For this reason, Kohler (1938: 104) and Metzger 1941: 14) have distinguished the phenomenal from the “trans-phenomenal” or “trans-experiential” world. Accordingly, if “physical world” denotes the col­lection of the results of physics and physiology, which can also account for the stimulation conditions and the mechanisms underlying perception, such a distinction does not imply any ontological dualism (Bischof, 1966b: 30f.). It is rather the restriction of the study of perception to what is theoretically admit­ted on the basis of physical and physiological knowledge, which is employed in designing the stimulation in experimental conditions, that brings a dualistic stance to the account, because it leads one to treat as mental or psychic subjec­tive data that which is not completely reduced to the stimulus property.

Source: Calì Carmelo (2017), Phenomenology of Perception: Theories and Experimental Evidence, Brill.