IX RETROSPECT. THE USE AND MISUSE OF ABSTRACTION I OOKING back over the theories we have considered, ^we see that for the most part they fall into one or other of two main types. One group of theorists lays stress upon the unity of the developing organism, though not disdaining analysis; the other tends to ignore the unity of the organism and deals in terms of separate or separable characters and their presumed material representatives in the germ. The difference, though not an absolute one, is deep-rooted, and can be traced right back to the time of the Greeks, where the contrast comes out very clearly in the opposing views of Aristotle and Hippocrates ; historical research would probably reveal it to be older still. Delage in his masterly survey of theories of development and heredity has already recognized a distinction between the ‘organicists’, as he calls them, and the ‘micromerists’. The distinction which I wish to draw corresponds roughly with his, but as I look at the matter from a rather different angle, I propose to group the theories as unity theories and -particulate theories.
This primary classification does not, however, sufficiently allow for the diversity of standpoint shown within the unity group, a diversity which philosophically may go very deep — as between Aristotle and Loeb, for instance. I have accord
ingly subdivided this group into the organismal and the physico-chemical. The organismal group of unity theories I have so named because they express in various degrees a view which we shall develop in the next chapter as the ‘organismal theory’ of living things. The physico-chemical group differ from the organismalists in accepting a mechanistic theory of the living thing, but they still hold to a unitary view of the organism, regarding it as a physico-chemical system. Their method is essentially physiological.
Along these lines, we can classify in the following manner the main points of view considered.
THE USE AND MISUSE OF ABSTRACTION 133
One or two theories do not fit comfortably into either of the two main categories. Roux’s theory, for example, is nearly, but not quite, a unity theory, and he certainly accepts in a measure the particulate conception of development. The same remark applies with even greater force to O. Hertwig, whose views we have not discussed for the very reason that he is essentially an eclectic, though with a heavy bias towards the particulate view. Another line of cleavage, which in practice, though not necessarily in theory, coincides approximately with the line we have drawn between the unity theories and the particulate theories, is that between the epigenesists and the preformationists, between those who hold that in development there is new creation of form and activity and those who regard development as a mere unfolding of what is there already, a making visible of an invisible complexity. In its fully developed form the preformationist view is always particulate, postulating a pointto-point correspondence between the germinal organization and the structure of the adult. The unity theory, on the other hand, is combined invariably with an epigenetic view of development. As we saw above (p. 29) in discussing the early preformationists and epigenesists, the one group is predominantly static and morphological in outlook, the other dynamic and physiological. Usually the preformationist
134 RETROSPECT. THE USE AND
adheres to a materialistic conception of life; the epigenesist
is often, but not always, a ‘vitalist’.
It is not proposed at this stage to give a full analysis of the two opposing views; their characteristics will emerge as the discussion proceeds. A short recapitulation, however, of the main points brought out in the preceding chapters will be of service in orienting us with respect tq them.
2. Aristotle is of course the father of the unity theory, and certain essential points are stated by him clearly and adequately. Working out his views in direct opposition to the Hippocratic theory of pangenesis, which is a typical particulate theory, he sets out the alternatives in masterly fashion. There is, in Aristotle’s view, one fundamental cause of the resemblance of parent and offspring which includes the point-to -point correspondence ; this is the realization of the same total functional potentialities. Hereditary resemblance is shown essentially in the functional activity of the organism as a whole, and the resemblance in minor detail is, as it were, derivative and incidental. Hereditary functional likeness is a consequence of similarity of development — of the realization of the same unified functional potentialities. Creative synthesis is then the primary cause alike of development and of heredity. It is this that brings about that ‘composition’ or wholeness of the body and its functions, that organic architecture and functional harmony, which make of it not an assemblage or composite of parts, but one unitary whole. The opposing particulate theory cannot account for this primary and fundamental unity of the developing organism. The semen can come only from the homogeneous parts or their elements, and not from the composition or arrangement which makes of them organs and an organism. There is therefore some primary cause of the total unitary resemblance which eludes the analytic method of the pangenesists. Development and heredity are then functions of the whole organism and are not fully explicable in terms of the activities of its parts.
Modern particulate theories attempt to account for this primary unity by assuming an architectonic complexity of
MISUSE OF ABSTRACTION 135
the germ-plasm, a unitary and organized structure, which corresponds in more or less detail with the structure of the developed whole, and is the ground of the unity and functional harmony of the developing organism. But this is merely to restate, in terms of a purely hypothetical material configuration, Aristotle’s contention that there is one fundamental cause of the unitary functional resemblance between parent and offspring. For Aristotle’s unified functional potentiality Weismann for instance substitutes the complex and organized structure of the Id, putting material preformation in place of functional potentiality. He recognizes therefore the existence of unity, but as something static.
Aristotle points out a further difficulty of the particulate theory, by remarking that if the ‘seeds’ (or representative particles) are different from the parts from which they come and which they make, 1 then the problem of development exists for each of the seeds, and the development of the whole is not explained by thus fragmenting it into a multiplicity of separate developmental processes.
The idea of development as a unitary process, which cannot be completely explained by analysing it into parts and stages, recurs with full force in K. E. von Baer, who maintains that one stage of development is not the sufficient cause of the next, but that all are ruled and bound up in a rhythm by the essential nature (die Wesenheit) of the organism to be. Contrast with this the guiding principle of W. His, some decades later, that each stage is the necessary resultant of the foregoing, and the sufficient cause of the ensuing stage.
Of the preformationists and epigenesists of the seventeenth and eighteenth centuries none is important methodologically save Bonnet, who very definitely foreshadows the Weismannian view. Darwin in 1868 revived the Hippocratic doctrine of pangenesis on a cellular basis, and Galton, by postulating a reserve stock of pangens in the germ-cells (the stirp), which is handed on from generation to generation, hit on the essential idea of the continuity of the germ-plasm.
1 Cf. Bonnet (p. 30) and Weismann (p. 50).
136 RETROSPECT. THE USE AND
The particulate view comes to full expression in Weismann. Here we have heredity explained as being due to identity, continuity, of germinal substance, and development regarded as the work of a particulate germinal organization present at the outset. All potentialities, however distant and eventual, are materialized as determinants or groups of such, quite irrespective of the occurrence of the outer events which may call them into play. The germ-plasm contains within itself material provision, in the shape of latent determinants, for all likely and even unlikely situations that may arise. One might almost say that Weismann puts the environment into the germ-plasm under the form of materialized potentialities representing eventual responses to all sorts of environmental conditions that may be encountered.
The futility of this mode of reasoning has been fully exposed by Delage, who asserts with reason that a latent character is an absent character and its hypothetical bearer or determinant a useless fiction. The fact that certain characters are separately heritable, Delage argues, does not necessarily lead to the assumption that they are represented by special determinant units in the germ, and we can get on perfectly well with a simple, non-representative ‘germplasm’, or better still with the egg-cell itself, regarded as a living organism, and as the earliest stage of the organism to be.
Development, in Delage’s view, is a real formation of new organs not present in any shape before, a real differentiating out of the generalized functions of the ovum, and in this process the environment plays just as important a part as the initial germinal constitution, which must be precise and specific, but need not be complex. Delage thus approaches from a different angle the simplicity of Aristotle’s conception of the unitary functional potentiality of the germ.
The physico-chemical group as a whole unite with Delage in condemning the particulate theory of development, as being otiose and misleading, and substitute the conception of the developing organism as a unitary physico-chemical system . 1 They recognize and accept accordingly the unity
1 See also below, pp. 141-2.
MISUSE OF ABSTRACTION 137
of the organism, as does also in large measure W. Roux. Very important is their and Roux’s insistence on the physiological process of metabolism as the foundation both of growth and of development.
The gene theory is primarily a theory of biparental inheritance, and throws little light on development. As an explanation of development it is logically less adequate than the Weismannian theory, since with its simplified germplasm composed of a linear arrangement of genes it tacitly ignores the problem of ‘composition’ and unity which Weismann attempted to meet with his architectonic germ-plasm. It tends to treat the organism as a bundle of separable characters — as a congeries, not as an organized whole.
With the mnemic theories of Hering, Butler, and Semon, there is introduced a new conception into the theory of development, that of the revival of past experience under the form of habit. Two quite distinct formulations of the idea are given, by Hering and Semon on the one hand, and by Butler on the other. The first two treat the phenomena from the materialistic standpoint, and account for mnemic phenomena on the supposition that the original stimulus leaves a material trace or engram in the substance of the organism. In Semon’s hands the mnemic theory of development tends to become a particulate theory, with engrams in place of determinants. Butler on the other hand interprets the facts from the point of view of psychobiology, and accepts fully the organismal conception. Development and reproduction are for him functions of the organism as a whole ; he makes no attempt to analyse them in terms of the activities of cellular and subcellular units. He holds that there is actual continuity of experience, not only within the bounds of the individual existence, but extending right back to the very dawn of life. For the material continuity of the germ-plasm he substitutes continuity of experience. In many ways his view of development is similar to Aristotle’s, but he makes the great advance of realizing the historical continuity of life from its earliest beginnings. Aristotle had no idea of this immense historical background to the development of the
138 RETROSPECT. THE USE AND
individual. In his view development makes a fresh start in each generation; the offspring is the child of its parents and not of its ancestors, and life is born afresh at every conception.
We have seen how this idea of the historical basis of organic activities opens up the possibility of explaining many of the principal features of development — its routine orderliness, for example, recapitulation, and anticipatory action in general.
3. We can best elucidate the contrast between the particulate and the unity theories and estimate their relative value as applied to the problems of development, if we consider the methods of abstraction which they respectively employ. All scientific theories necessarily abstract to a greater or less degree from the realities studied, and the character of a theory depends as much upon the mode of abstraction employed as upon the nature of the facts on which it is based. If we consider the various ways in which, for the purposes of science, abstraction is made from the living reality of the organism, we shall see how the different theories of development have arisen, and how their character has been determined by the mode of abstraction they adopt.
Biology occupies a unique and privileged position among the sciences in that its object, the living organism, is known to us not only objectively through sensory perception, but also in one case directly, as the subject of immediate experience. It is therefore possible, in this special case of one’s own personal life, to take an inside view of a living organism. Naturally the direct intuitive understanding obtained by this immediate experience of living cannot form the subjectmatter of biology, which is an objective science, and we must be very chary of reading into the activities of other organisms the motives and modes of experience which we discover in ourselves. Nevertheless introspective knowledge does give us an insight into the reality of the living organism which cannot be otherwise obtained, and supplies us with a standard by which to test our conceptions of the living thing.
When we conceptualize this living experience, we arrive at a definition of organism which, though it is abstract and
MISUSE OF ABSTRACTION 139
schematic as compared with the experienced reality, is yet rich in content as compared with the still more schematic representations commonly employed in biology. The concept of organism which we derive from a study of direct experience is that of a continuing psycho-physical unity or individuality, which acts as a whole in relation to its environment. We shall develop his ‘organismal’ view of the living thing in the next chapter, as the basis of our theory of development and reproduction. We shall there see that for the purposes of organismal biology it is necessary to expand this definition of organism somewhat, to bring in as essential factors the past and future of the organism and the relation of its activities to the three main ends of life — development, maintenance, and reproduction. For our present purpose however, which is to trace the steps by which for this concrete concept there come to be substituted various more abstract and restricted scientific concepts, it is sufficient to deal with the simple definition of organism, as a psychophysical or organic unity persisting in time and acting as a whole in relation to its surroundings. ‘Psycho-physical’ is here used to denote that the activity of the organism is of a different order from purely physical activity. This point will be further elucidated later (pp. 172, 183-5).
4. We may distinguish two main modes of abstraction which are in practice often combined. In the first, abstraction is made from the organism as a whole, and the idea of its unity is in some measure retained. Such is the form of abstraction employed in the unity theories, and according to the degree of abstraction practised they fall into the organismal or the physico-chemical group. In the second, the mode of abstraction is analytical, and resolves the organism into an assemblage of more or less abstract parts and separate processes. This is the method which leads to the particulate view of development.
We shall deal first with abstraction applied to the organism as a whole.
As we have already seen, the first stage of abstraction from the concrete reality of organism as experienced is the concept
140 RETROSPECT. THE USE AND
of organism as a psycho-physical unity or individuality. If the process of abstraction is stayed here, we get the organismal view — the view which we shall develop in the next chapter. Almost invariably the process is carried farther, and for the purposes of biology abstraction is made of the psychical aspect of living things. Our first group, however — Aristotle, Wolff, Baer, and Butler — hold the full organismal position, avoiding this abstraction of the psychical. This appears also to be the theoretical position of Conklin, though it is hardly maintained by him fully in practice. It is also the view explicitly held by Ritter and some others, whose general views we shall consider later (pp. 176 et seq.).
The second step along the path of abstraction — namely, the elimination of the psychical — is one which nowadays is almost universally taken as a matter of course. It is a step of immense importance, for it introduces at once a dualism of matter and mind, and creates between them a dividing line which can never be crossed. To reintroduce mind into living things, to reconstitute the living unity, it is necessary to have recourse to such lame expedients as psycho-physical parallelism or psycho-physical interaction, or to adopt some form of dualistic vitalism. With the psychical aspect eliminated, the organism becomes a material system, similar in nature to, though more complex in structure than, other material bodies.
The complexity and variety of organization naturally provoke investigation, and give rise on the one hand to the science of organic form, in which types of structure are distinguished and their variants classified (morphology), and on the other hand to the study of the functioning of the different types (physiology). Morphology tends to remain a formal and abstract science, until it is revivified by the study of function; physiology develops very soon the concept of the organism as a complicated mechanism. Here two somewhat diverse points of view emerge — the teleological and the dynamical.
A machine is definitely a teleological construction, and the working of its parts can be fully understood only if their
MISUSE OF ABSTRACTION 141
relation to one another and to the action of the whole be realized and grasped. The same view can be applied to the organism, which may be regarded as a teleological mechanism or machine, albeit of extreme complexity. The teleological point of view has undoubtedly great heuristic value in biology, and is in fact much used.
But the progress of physico-chemical study applied to the living thing has shown that the organism cannot be separated from its environment, with which it maintains the closest relations of interchange of matter and energy. Hence the conception arises of the organism as a physico-chemical system, standing in closest connexion with its physicochemical environment. The simple concept of the organism as a formed machine is then replaced by the more general concept of it as a dynamical system. From this to the application to the organism of the general philosophical concept of material determinism there is only a step, and the organism tends then to become merged in, and hardly distinguishable from, the general flux of material events. We shall consider later the consequences of a purely materialistic treatment of living things.
It is to be noted that in both the teleological and the biochemical formulations the organism retains a considerable measure of unity — very definitely so when it is considered as a machine, and to a considerable extent also when it is regarded as a dynamical system. But both formulations are abstract as compared with the living reality of organism, or even with the primary concept of it as psycho-physical unity or individuality.
In our survey of theories we have not encountered any pure machine-theory of development, and this is easily comprehensible, for the machine-concept is primarily applicable to the formed organism, and only with great difficulty to the organism in process of formation. A machine that constructs and repairs itself is something beyond our present ken.
Of theories treating the organism in formation as a physicochemical system we have seen several — those of Loeb and Child for example, and in considerable measure also those of
14 * RETROSPECT. THE USE AND
Delage and Verworn. Loeb finds the ground of the harmonious structure and unitary functioning of the organism in the fact that it is from the very beginning in the egg a unified physico-chemical system. Child, who fully realizes the fact of unity or wholeness, regards the developing organism as a reaction-system behaving as a whole. He takes the proper physiological view that structure or organization is not a cause but an effect of metabolic functioning, that the organism is a dynamical rather than a statical unity. Modern physiology teaches that the cell is a chemo-dynamic system, which requires constant oxidation to preserve its structure. ‘The life machine is therefore totally unlike our ordinary mechanical machines. Its structure and organization are not static. They are in reality dynamic equilibria, which depend on oxidation for their very existence.’ 1
Delage’s ultimate goal is the explanation of development in terms of its physico-chemical determinism, but he treats the organism and the cell as at least provisional unities. The importance of the cell as a natural biological unit is also stressed by Verworn.
All these physiological theories then, though abstracting completely from the psychical aspect of living things, retain in some measure the concept of the organism as a whole in regarding it as a physico-chemical system. They properly belong therefore to the unity group.
Roux’s position is an interesting one. He comes very near the organismal view by his insistence on the functional activity of the living organism, and the need to state the laws of biology first in terms of elementary functions. He is, however, prevented from regarding the living thing as an organic or psycho-physical unity by his acceptance of the materialistic conception, with its elimination of the psychical. The functions or ‘complex components’, in terms of which he attempts to explain development, can therefore never be completely joined up to a central unity or organism in the concrete sense. Nevertheless his attempt to work with
1 F. G. Donnan, ‘The Mystery of Life’ (Brit. Assoc. Address), Nature , vol. cxxii, 1928, p. 514.
MISUSE OF ABSTRACTION 143
biological rather than physico-chemical concepts is highly significant, and puts him ahead of the other physicochemical theorists.
Turning back now to the morphological method of study, we see at once that it is by its very nature an abstract method. The morphologist concentrates on the structural aspect of living things, and regards function as derivative from structure. Contrast the physiological view which we have outlined above. Applied to development the morphological method of thought may be characterized as the kinematic, since it sees in development merely a succession of structural phases, each of which is the resultant of the last, and the cause of the ensuing, phase (cf. His, p. 95). Linked with this static or kinematic mode of abstraction, which betrays the morphological habit of thought, there is usually a strong tendency to disregard or abstract from the environment as a developmental factor. The morphological abstraction in its entirety exerts its chief influence on the theory of development when combined, as it usually is, with the analytical mode of abstraction which we shall shortly proceed to discuss. The weakness of the morphological approach to the problems of development and heredity is well brought out by D’Arcy Thompson in the following passage.
‘It would seem evident’, he writes, ‘that, except in relation to a dynamical investigation, the mere study of cell structure has but little value of its own. That a given cell, an ovum for instance, contains this or that visible substance or structure, germinal vesicle or germinal spot, chromatin or achromatin, chromosomes or centrosomes, obviously gives no explanation of the activities of the cell. And in all such hypotheses as that of “pangenesis”, in all the theories which attribute specific properties to micellae, idioplasts, ids, or other constituent particles of protoplasm or of the cell, we are apt to fall into the error of attributing to matter what is due to energy and is manifested in force: or, more strictly speaking, of attributing to material particles individually what is due to the energy of their collocation. . . . The things which we see in the cell are less important than the actions which we recognize in the cell.* 1
1 On Growth and Form , Cambridge, 19 17, pp. 157-8. Cf. Child’s dynamic view of organization.
144 RETROSPECT. THE USE AND
Abstraction from the living reality of organism reaches of course its highest degree in the mathematical or statistical treatment of living things. Here the organism is regarded merely as a numerical value — a number, a weight, a dimension. Mathematical laws of growth, for instance, may be worked out, in which the organism is treated simply as a quantity which increases in accordance with a certain formula. Clearly such formulation gives only the most general and abstract account of the process, highly useful though it may be within strict limits.
To recapitulate the main stages of abstraction from the organism as a whole — we get from the living reality as experienced, which is our ultimate standard, first, the primary abstraction or conceptualization as psycho-physical unity or individuality, from which may be developed the organismal theory of living things, and second, by abstraction from this of the psychical aspect, the ordinary ‘scientific’ conception of the organism as a machine, or more generally a physicochemical system. This gives us the mechanistic abstraction, which may remain however to some degree integrative, i. e. may still regard the organism as to some extent a unity. According as stress is laid on form or on functional activity, we get either a static-kinematical or a dynamical conception of the organism.
5. The use of analysis is characteristic of science generally. Given a complex body, the chemist proceeds immediately to resolve it into its elements, to determine their relative proportions, and in some cases their architectonic arrangement. The same tendency is shown very clearly in biology. Given an organism, the morphologist’s first thought is to discover its structure in minutest detail, to resolve it into its constituent organs and cells and their arrangement. The same process of analysis is applied to what appears to be the ultimate vital unit, the cell; this is decomposed into its constituent parts, nucleus, cytoplasm, chromosomes, mitochondria, and so on, and each of these elements is further resolved as far as may be into smaller parts, as for instance chromomeres, linin threads, and granules of all kinds. The
MISUSE OF ABSTRACTION 145
process is even extended beyond the limits of microscopical vision, and hypothetical units, such as biophors, bioblasts, and genes, are freely invented to fill the gap between the just visible units and the complex colloidal molecules which make up the bulk of living matter. The method of morphological analysis leads then to a biological atomism, analogous to the atomism of the chemist.
From the physiological side also there starts a similar process of analysis or decomposition. The physiologist studies for choice isolated organs or organ-systems — the properties of a muscle-nerve preparation, for example, or the functions of the isolated heart. Even when he studies a major organsystem as a whole, the nervous system for instance, his method remains analytical; he resolves the action of the nervous system into the action of reflex arcs in their interconnexion with one another.
Another analytical method of study is to concentrate on cell-physiology, in the hope of explaining the activity of the organism as a whole in terms of the functions of its presumed ultimate units or constituents. To quote Donnan again: ‘We must decompose or analyse the great mass phenomena of life into their elementary unit or constituent phenomena. To-day general physiology in its application of physics, chemistry, and physical chemistry to the living cell, is the fundamental science of life 5 (ibid., p. 513).
Now this analytic method, employed both in the study of form and in the study of function, is quite indispensable in biological research, and has yielded extremely valuable results. It is essential also for organismal biology. But we must note that it necessarily entails abstraction. The initial step which leads to abstractness of treatment is of course the isolation and definition of parts and part-processes as such. To define is to separate, and to separate is to ignore or to disregard in some measure the relations with other parts and with the whole. In the living thing there are in actuality no separate parts, no separate processes, for no part can be adequately characterized save in terms of its relations to the whole.
By the process of morphological analysis we can, for instance,
i 4 6 retrospect, the use and
resolve the organism into its component cells, but the cells so distinguished are abstract morphological units, characterized statically, in terms of structure. Actually the living tissue-cell is indissolubly linked up, by reason of its functional activity, with the neighbouring cells, and, through the milieu interne and the nervous system, with the general activities which the whole organism is pursuing. The tissue-cell takes part in the activity of the whole, and it is dependent for its own continued existence as a living part upon its manifold functional relations with the whole. If we distinguish it as an independent unit or component we necessarily abstract from its full reality; we disregard its functional connexions or relations with the whole, and form a simplified and static conception of it.
In the living thing there are no completely separable or independent parts ; if we distinguish separate units or components it is at the cost of artificially simplifying our definition of them by abstracting from their continuing relations with the activity of the organism as a whole. It is primarily because the parts or constituents so distinguished are to a large extent abstract that it is impossible fully to reconstitute from them the whole from which they are themselves derived by the process of analytical abstraction. This is true even if we characterize them physiologically.
Contrast in this respect a machine. The machine has separate parts; it can be taken to pieces and put together again; its parts can be adequately characterized in terms of their own structure, apart from their relations to the machine as a whole. This is not the case with the organism. Here the parts can be adequately characterized only in terms of their functional relations to the organism as a whole. These relations, which are manifold and subtle, involve time and process, a taking part or merging in the total activity of the continuing unity which is organism.
The unity of the organism is accordingly not decomposable without loss, and cannot be resynthesized in its original completeness from the abstract components distinguished by analysis. We may sum this up in the following cardinal law
MISUSE OF ABSTRACTION 147
of biological method: The activity of the whole cannot be fully explained in terms of the activities of the parts isolated by analysis , and it can be the less explained the more abstract are the parts distinguished.
Since analysis is necessary for biological science we must accept the fact that our biological results will be to a certain extent abstract and schematic, and we must strive to correct this abstractness as far as possible by distinguishing only such elements as are concrete and biological, not physicochemical and abstract, and by carrying out as complete a reconstitution or reintegration of such elements as may be possible.
It follows from what we have said that the parts cannot be understood save in relation to the whole, and so we arrive at our second law of biological method : No part of any living unity and no single process of any complex organic activity can be fully understood in isolation from the structure and activities of the organism as a whole. To regard any process or structure by itself without relating it to the general activity of the organism is to deal with something which is in large measure abstract and unreal. To re-invest it with some degree of concrete reality it is necessary to re-integrate it into the whole. Its isolation by analysis should be provisional only, and after analysis there should always follow re-integration. We know that the reconstitution of the original unity will be incomplete, but we must make it as complete as possible.
The necessity for envisaging every physiological process in its relation to the activity of the organism as a whole has been fully stressed by Dr. J. S. Haldane and his followers (see below, pp. 174-6), who take up a position which is in practice, if not in theory, similar to the organismal. Even from the mechanistic standpoint, the need is recognized, by both morphologist and physiologist, for resynthesis of the parts and part-processes isolated by analysis. Both would agree that the meaning or significance of the parts and processes they isolate from the whole can be fully realized only if they are put back again in their place and seen in relation to the structure and the activity of the whole.
148 RETROSPECT. THE USE AND
Clearly no part or organ lives a completely independent life, but can exist and perform its proper function only in the closest dependence on, and co-ordination with, the rest of the body. All parts and organs are, as Kant would say, reciprocally means and ends, and all co-operate in the life of the organism as a whole.
‘For’, as D’Arcy Thompson writes, ‘the life of the body is more than the sum of the properties of the cells of which it is composed: as Goethe said, “Das Lebendige ist zwar in Elemente zerlegt, aber man kann es aus diesen nicht wieder zusammenstellen und beleben” * (ibid., p. 38).
I cannot forbear to quote another passage from the same author where the importance of the unitary or integrative view is most clearly brought out.
‘We tend, as we analyse a thing into its parts or into its properties, to magnify these, to exaggerate their apparent independence, and to hide from ourselves (at least for a time) the essential integrity and individuality of the composite whole. We divide the body into its organs, the skeleton into its bones, as in very much the same fashion we make a subjective analysis of the mind, according to the teachings of psychology, into component factors : but we know very well that judgment and knowledge, courage or gentleness, love or fear, have no separate existence, but are somehow mere manifestations, or imaginary co-efficients, of a most complex integral’ (p. 712).
So it is with the organs of the body — they form one harmonious and indissoluble whole.
‘Muscle and bone, for instance, are inseparably associated and connected; they are moulded one with another; they come into being together, and act and react together. We may study them apart, but it is as a concession to our weakness and to the narrow outlook of our minds. We see, dimly perhaps, but yet with all the assurance of conviction, that between muscle and bone there can be no change in the one but it is correlated with changes in the other; that through and through they are linked in indissoluble association: that they are only separate entities in this limited and subordinate sense, that they ar t farts of a whole which, when it loses its composite integrity, ceases to exist. The biologist, as well as the philosopher, learns to recognize that the whole is not merely the sum of its parts. It is this, and much more than this. For it is not a bundle of parts, but an organization of
MISUSE OF ABSTRACTION 149
parts, of parts in their mutual arrangement, fitting with one another, in what Aristotle calls “a single and indivisible principle of unity” 9
(p P . 713-14) That the parts can be understood only through their relations to the whole is then a cardinal principle of all sound biological thinking, and it is recognized as such by all competent thinkers.
In the foregoing discussion I have for the sake of contrast somewhat over-exaggerated the difference between the unity theories and the particulate. There is in practice no absolute antithesis between them. Even on the particulate conception the unity of the organism cannot be completely ignored, and we have seen how Weismann for example implicitly recognized it in his concept of the Id as an organized unit. No one would nowadays seriously maintain that the organism is merely a bundle or assemblage of unit characters; the existence of ‘composition’, arrangement, organization, unity, can simply not be disregarded. On the other hand, even the organismalist, though firmly convinced that the unity of the living thing cannot be completely explained in terms of the activity of its parts, will not hesitate to make use of the method of analysis in the practical conduct of his researches. He will employ the method of analytical abstraction where necessary, remaining conscious of its limitations, and seeking always to reintegrate into the activity of the whole the partial activities which he distinguishes.
This very difficult but fundamental question of the relation of the parts to the whole, which I have treated rather inadequately and abstractly above, will be further considered in the next chapter, where an attempt is made to work out a more adequate concept of organism, taking account of the all-important temporal relations of organic activity. The life of an organism is essentially a unitary functional or dynamical process, in which ‘whole’ and ‘parts’ are inextricably interconnected. Both whole and parts are together the expression of the life of the individual.
6. So far we have considered the legitimate use of the analytical method, the inevitable abstractness of which can
ISO RETROSPECT. THE USE AND
be to some extent mitigated by reintegration. There is, however, a misuse of the analytical or disintegrative method which leads to disastrous consequences, many of which we have seen in our survey of theories of development.
There is, for example, a tendency on the part of some authors to neglect the natural biological unities of organism and cell and to speak of ‘living substance’ as if it were a selfexistent thing. Actually life, so far as observation goes, exists only in the form of individual units, and ‘living substance’ is therefore a pure abstraction, to which nothing corresponds in Nature. The process of abstraction, of elimination, is here carried too far; essential elements are left out, and the result is a purely fictional concept. A further misuse of this already abstract concept is to speak of a ‘germinal substance’, as Weismann and so many others do. It is true that Weismann means by this something more than mere substance, namely the complex architecture of the Id, but even in this form his germinal substance is something quite abstract and unsubstantial.
More dangerous still is the misuse of the method of morphological analysis. The organism is by this method resolved into cells, cells into their constituent parts, and the substance of the cell into hypothetical units, to which are attributed many of the essential vital functions. This fractionalization is a method of approach to the problems of heredity and development which has become traditional and habitual, so that nowadays any other way of looking at these problems is rarely considered, and it is of course the basal method underlying all particulate theories. It is generally, though not invariably, coupled with the idea that some at least of these ultimate units represent and give rise to, or at the least co-operate in the formation of, particular parts or characters of the organism. This idea of representative particles is, we have seen, a very old one, dating back at least to the Greeks, and revived again by Bonnet, Darwin, and Weismann. It derives some of its force and verisimilitude from the fact that certain characters appear to behave as units in inheritance — a particular lock of white hair, for
MISUSE OF ABSTRACTION 151
example, mav recur from one generation to another. From facts of this kind it is easy, but illogical, to conclude that all characters of the organism are separable in inheritance, that the organism is, as it were, a bundle of separate characters, represented separately in the germ, which can be shuffled about, so that some of the offspring get one set, some another, and so on indefinitely. 1 It may be remarked that to distinguish separate characters at all in the organism has necessarily something artificial and abstract about it (see above, p. 77). Obviously the number of characters that can be distinguished is infinite, but yet none of them is in reality separate from the rest. The lock of hair, for instance, clearly cannot arise apart from the organism which manifests it. Separate or separable characters are therefore to a very large extent abstractions. But the idea that the organism is a composite of separate characters, each of which is represented in the germ by a separate vital unit, seems to have a perennial fascination for the human mind.
There are two points to be distinguished in considering the resolution of the cell into ultimate vital units. The first is that the procedure is logical enough so far as the actual existence of discrete units is concerned. Microscopical analysis reveals in the cell the existence of smaller and smaller discrete units; there is no reason to believe that the limits of microscopic or ultra-microscopic vision coincide with the limits of discreteness of living matter; physical chemistry leads us to think that the ultimate basis of living protoplasm is a colloidal solution in which are found dispersed
1 Cf. C. E. McClung: ‘If the germ cells are, then, a register of racial experience, it is clear that a mechanism for assorting and recombining the pages of this story would tend to an ultimate uniformity of expression with almost all degrees of individual variation. Such possibilities seem to lie within the nature of the chromosome behavior in maturation and fertilisation. The haploid group of either parent cell contains the full register of class characteristics, and, upon fertilisation, the individual is provided with a double record. In its germ cells the double pages are separated and assorted by chance into different units, so that, if there are but eighteen of these pages in the single series, the possible varieties of it are 262,144. When later fertilisation combines these in duplicate, the possible expressions of variation reach the incomprehensible number of 68,719,736/ In Cowdry, General Cytology , Chicago,
152 RETROSPECT. THE USE AND
particles of molecular size. Hence it is quite possible, and even likely, that ultramicroscopic units exist, intermediate in size between visible granules and colloidal particles. Briicke and Herbert Spencer long ago arrived at this conclusion on theoretical grounds. The conception of the cell as built up of ultimate and invisible living units is widespread among modern biologists . 1 O. Her twig resolves the cell into bioblasts, M. Heidenhain into protomers, Weismann into biophors, de Vries into pangens, and so on. Hertwig’s bioblasts, for example, are the ultimate vital units that possess the power of assimilation, growth, and division. They cannot be further split up without losing these distinctively vital properties.
‘Just as plants and animals can be resolved into thousands and thousands of cells,’ writes Hertwig , 2 ‘so also the cell itself is built up of numerous elementary vital units, which lie below the limits of microscopical visibility, are different from one another chemically, and form here the cytoplasm and its numerous differentiation-products, there the nucleus, the nuclear membrane, the linin threads, the chromosomes, the trophoplasts and so on, and as integral parts of an organism stand in organic relations with one another. “As physics and chemistry go back to molecules and atoms, so must the biological sciences penetrate down to these units, in order to explain from their inter-relations the phenomena of the organic world” (De Vries).’
So far, so good ; these hypothetical units are arrived at by an extension of the legitimate method of analysis that led to the formulation of the cell-theory; the procedure is logical, though it leads to further abstractness.
The real mischief begins when these hypothetical units are regarded as the sole hereditary elements, and as the formative agents in development. The train of reasoning which leads to this conclusion runs somewhat as follows. It is axiomatic, the theorists assert, that the male and the female contribute equally to the heritage of the offspring. The only equivalent structure in the male and the female gamete is the nucleus. The nucleus must therefore carry all
1 It does not however find favour with modern physiologists, cf. F. G. Hopkins, Nature , xdi, 1913, p. 220.
2 Allgemeine Biologie , 6th and 7th edit,, Jena, 1923, p. 65,
MISUSE OF ABSTRACTION 153
the hereditary characters, and the nuclear substance must be the controlling and differentiating agent in development. Further, it must be of an extremely complex organization, since we must suppose it to contain the material representation of each separate character of the organism. If it contains particles representative of all the characters, these must be of ultra-microscopic size, else there would not be room for them in such a minute object as the chromatin head of a spermatozoon. We are led therefore to the conception of a complex germinal or nuclear architecture built up of ultramicroscopical units — the biophors of Weismann, the pangens of de Vries, or what other names may suit the fancy. They correspond in fact, though not necessarily in ascribed function, to the physiological units distinguished by structural analysis pushed to its theoretical limit.
This conception of an architectonic and formative germplasm is of course most fully developed by Weismann, and has been considerably modified and softened down in more recent days. Weismann, however, with de Vries (as well as those who believe in the real existence of genes as particulate bodies), does undoubtedly attribute to these ultimate vital units amazing powers of determining and differentiating the characters of the developing organism. We have already pointed out in dealing with Weismann’s theory that he ascribes to the action of his biophors and determinants, which are ‘parts’ of the cell, effects which in reality result from the functional activity of the cell or the organism as a whole. There is introduced therefore a curious dualism or opposition between the active formative agent and the matter acted upon. This dualism is already evident in Darwin’s theory of pangenesis, where the germ-cell is to all intents inert until activated by the entrance of the pangens — it is the soil waiting for the seed, Aristotle’s germinal ‘material’ waiting for the sensitive soul to be supplied by the semen of the male. In de Vries’ theory, also, just as in Weismann’s, the cytoplasm is regarded as material to be acted upon and transformed by the pangens or determinants, issuing from the nucleus like a swarm of workers intent on
154 RETROSPECT. THE USE AND
their task. Here we re-encounter that analogy between the germ-plasm considered as an active agent, controlling the metabolic processes of the cell, and the immaterial entelechy of Driesch, upon which we have already remarked (see above, pp. 50 and 102). The germ-plasm is, as it were, a material entelechy. The attempt to find an internal formative mechanism as the cause alike of heredity and development, which is characteristic of nearly all modern theories, results necessarily in this separation of agent and material, just as the attempt of the vitalists to reintroduce life into the mechanistic abstraction that stands for organism results in a dualism or opposition between the immaterial agent and the material mechanism which it in some way controls. In either case one arrives at a Deus in machina. 1
The nuclear organization, the germ-plasm, or the genecomplex of modern theories, is accordingly invested with semi-magical powers of control. To show that this is no exaggerated statement, let me quote the words of a recent author: ‘Taken together, the chromosomes represent the sum total of all the elements of control over the processes of metabolism, irritability, contractility, reproduction, &c., that are involved in the life of an organism, but in the measure that organisms differ, so do the natures of their controlling mechanisms.’ 2
How is this truly astonishing conclusion arrived at ? By misuse of the method of analytical abstraction. We note first of all that the point of view is essentially morphological. In dealing with development abstraction is made of function and environment; the ordinary physiological functions of the cell and the organism are left out of the picture entirely.
The organism is regarded as a collocation of subordinate parts, of units of diverse degree, but the problem of ‘composition’, organization, or wholeness is ignored, and attention
1 One may call this curious inversion of the problem th t fallacy of reduplication , in order to indicate the fact that the functional manifestations of the whole are assumed to be due to an internal mechanism, or entelechy, or agent, which as it were doubles the role.
C. E. McClung, ‘The Chromosome Theory of Heredity*, in Cowdry, General Cytology , Chicago, 1924, p. 634.
MISUSE OF ABSTRACTION 155
is concentrated on the lowest grade of these units, the biophors, bioblasts, pangens, or what not. These are supposed to represent the parts or the characters of the developed organism, and in some way, which always remains mysterious, to give rise to them in the course of development. (The hereditary units being the purest of abstractions, it is of course natural that their relations with the characters they determine should remain obscure.)
Hereditary units and ‘determinants’ of all kinds are pure abstractions; the process of analysis has been carried so far that it is impossible to reconstitute from these purely abstract elements the activities of the cell or the organism as a whole. All that is left then to the theorists is to smuggle back into the determinants or other ‘parts’ the powers and functions which belong rightly to the organism as a whole, and have inevitably been dropped out during the process of analysis. The concept of the organism as a whole, which has been destroyed by unrestrained analysis, is reintroduced surreptitiously, and the qualities and powers of the organism as a whole attributed to certain abstract and subordinate farts of it, just as to entelechy are ascribed powers and capabilities which properly belong only to the whole organism.
The result of this method of approach is the unnecessary complication of an already sufficiently difficult problem. The particulate type of theory substitutes for the relatively simple visible facts of development an excessively complicated re-description of them in terms of an organization of purely abstract and fictional elements, which it locates within the nucleus — as it were, an invisible and active organism or agent within the visible organism. The inadmissibility of all particulate theories from the standpoint of a dynamical or physiological treatment of development has been fully and adequately demonstrated by Delage and Child, but to emphasize the point I may quote a particularly clear summing-up by J. A. Ryder . 1 His general standpoint is that
1 ‘A Dynamical Hypothesis of Inheritance’, Wood's Holl Biological Lectures for 1894, Boston, 1895, pp. 23-53.
156 RETROSPECT. THE USE AND
‘metabolism is . . . the sole agent in effecting the mechanical and dynamical rearrangement or sorting of the molecules into organs during development’. He writes:
‘Specially endowed corpuscles or “biophors” are not only needless as conditioning form or function, but also out of the question, dynamically considered. No creature can be supposed to have its life or germinal properties associated only with certain corpuscles within it, since we cannot suppose an organized whole dominated by a portion of it; it is not possible, for example, to conceive of individual life except from the entire organism that manifests it. There can be no “biophors” — bearers of life — the whole organism must do that as an indivisible unit’ (p. 52).
The physiological point of view could not be more clearly expressed.
The organismal attitude is similar, and has been stated by Ritter 1 as follows :
‘According to the organismal conception, all life phenomena, including those of inheritance, consist in the activities and interactivities of an enormous number of substances and units and forces, all of which, in exhaustive analysis, are dependent upon the organism as a living whole. It is, therefore, as futile to hunt in one corner as another for the physical basis of heredity in an exclusive and more or less metaphysical sense.’
7. In the above discussion the type of theory criticized is the pure particulate theory, which assumes in greater or less degree a point-to-point correspondence and causal connexion between the units of the germinal architecture and the parts or characters of the developed organism. More modern theories, such as the theory of the gene, are not quite so schematic and crude, and the idea of ‘determinant’ in the full Weismannian sense has now become practically obsolete. The supremacy of the nucleus is still fully maintained, but the chromosomes and their constituent elements, the genes, are regarded more as conditioning than as determining the characters of the organism; their action is differential rather than fully determinative. It is probable that the physiological interpretation of the gene hypothesis is not far
1 W. E. Ritter, The Unity of the Organism , Boston, 1919, ii, p. 33.
MISUSE OF ABSTRACTION 157
distant, when the gene will cease to be regarded as a selfexistent particulate unit and will be merged in the general physiological activity of the nucleus. But the gene theory in its original formulation, where genes are treated as real entities, is tainted with the same faults as the particulate theory of Weismann — the resolution of the organic unity into isolated parts or characters, the opposition of the nuclear substance as active and formative to the cytoplasm as comparatively inert material, the invention of purely abstract and Active units or parts endowed with mysterious formative powers. So long as the gene is regarded as a purely hypothetical concept invented for the specific purpose of explaining certain complicated facts of inheritance, no great harm is done. But when it is treated as a real existent body all the factitious and gratuitous complications of the particulate theory tend to set in. The hypothesis is accordingly one to be employed with the greatest circumspection.
In particular the idea of nuclear dominance, which it shares with the older particulate theories, appears abstract and non- physiological. From the point of view of physiology the relations between nucleus and cytoplasm are obviously complex, intimate, and ever-changing. There cannot be any absolute separation between the functions of the nucleus on the one hand and the functions of the cytoplasm on the other. Their relations are reciprocal, each affecting each in constant succession. Nor can cither be understood save in relation to the other, and to the activity of the cell as a whole, for neither is capable of long-continued existence apart from the rest of the cell. To establish then a rigid distinction between the nucleus and the cytoplasm, to allot to each element clearly defined and separate functions, is to deal with unreal abstractions. To regard one as controlling the other is quite illegitimate and introduces that dualism of agent and thing acted upon which runs through and vitiates all theories of nuclear dominance.
8. We have hitherto touched very lightly upon the philosophies underlying the various theories of development and heredity of which we have treated. We have found it
158 RETROSPECT. THE USE AND
more important to discuss the actual methods of attack upon the problems which have been adopted, and we have traced how by the use and misuse of abstraction certain of these theories get farther and farther away from the living reality.
It is very curious how the same types of theory recur throughout the ages. The contrast between the unity theory and the particulate is fully developed among the Greeks, who also, as ;s well known, foreshadowed some other ‘modern’ scientific conceptions, as for instance the theory of evolution. This would seem to indicate that certain types of explanation are natural to the human mind, and that the diversity shown arises not so much from the facts, with which the ancients were singularly ill-acquainted, as from some fundamental diversity of mind. Some men are organismalists by nature, others take naturally to the particulate view. It would seem that man creates science in his own image, and that different types of mind create different kinds of science.
There are also underlying this initial diversity of standpoint other differences arising from the particular philosophy consciously or unconsciously adopted. Aristotle, for instance, holds very special philosophical views of his own, which are quite different from those of modern times, and his constructive theory of development is, as we have seen, thoroughly imbued with his particular philosophy. The preformationists were believers in the principle of mechanism introduced in their time by Descartes and the early physicists, and they took refuge in the theory of preformation because a mechanistic explanation of epigenesis appeared to them impossible. The epigenesists did not accept this way out of the dilemma, and frankly embraced vitalistic views.
K. E. von Baer was a believer in a sort of universal vitalism, and was ready to regard matter and spirit as alike human abstractions from the living reality. He also was an epigenesist.
When we come to the modern theories we find that, with some few exceptions, they are based upon the materialistic conception of reality. In a book to which we shall often have occasion to refer in the succeeding chapter, Whitehead
MISUSE OF ABSTRACTION 159
enunciates the following valuable maxim: ‘When you are criticizing the philosophy of an epoch, do not chiefly direct your attention to those intellectual positions which its exponents feel it necessary explicitly to defend. There will be some fundamental assumptions which adherents of all the variant systems within the epoch unconsciously presuppose. Such assumptions appear so obvious that people do not know what they are assuming, because no other way of putting things has ever occurred to them.’ 1 This is particularly true of most modern theories of development, for they are based upon a materialistic philosophy which is by their authors accepted without question or inquiry, and as a matter of course. To most biologists, the existence of the germ-plasm, for example, appears axiomatic and unquestionable; actually the concept of germ-plasm is derivative from the materialistic philosophy, and has only this restricted and relative value.
The domination in modern times of a materialistic or mechanistic basal philosophy is obvious in the physicochemical or physiological theories; it is equally clear in the germ-plasm and gene theories, and W. Roux, though in practice getting somewhat beyond the materialistic method, yet bases his theoretical views directly upon the mechanistic philosophy, as being the only sound foundation of exact science. Butler is of course a definite exception, with his psycho-biological interpretation of development and evolution, but the other mnemists, Hering, Semon, and Francis Darwin, have not shaken free of the materialistic notion of physical traces or engrams, and remain wedded to a form of biological atomism.
The materialistic theory of reality is nowadays somewhat vieuxjeuy and has been very considerably shaken and modified by the new views of the constitution of matter and the establishment of the principle of relativity. It is, however, in its classical Laplacean form that it influenced and still influences biological method. This classical materialism may be defined as the belief that all physical phenomena, including the phenomena of life, are in ultimate analysis the
1 A. N. Whitehead, Science and the Modern World , Cambridge, 1926, p. 61.
160 RETROSPECT. THE USE AND
outcome of the motion of material particles, that such material particles are bound up in rigidly determined spatial systems or configurations, the successive stages of which follow one another in time according to immutable law, the position and acceleration of each particle being completely conditioned or determined by the immediately antecedent states of the changing configuration which is reality. The progression of events as perceived by the senses is merely the outward show of this incessant movement of invisible and ultimate particles.
It is a conception singularly ill-adapted to explain heredity and development, for as Picard writes, ‘Classical mechanics has led to types of differential relations, by postulating, more or less explicitly, a principle of non-heredity, according to which the infinitesimal changes that occur in a system depend solely upon the present state of that system.’ 1
It is under the influence of this mechanistic schematism, accepted as the true and only foundation of science, that the biologist proceeds at once and without hesitation to carry out that decisive abstraction of which we have spoken above — the elimination of the psychical. This done, he can treat the living thing as a complex mechanism — as a machine, or as a physico-chemical system, or, theoretically at least, as an artificially isolated section of the general configuration flux. In practice, we may remark in passing, he is often not consistent in his materialism, for he switches back on occasions to the primary organismal view, as when he ascribes to living things perceptions and feelings, or uses concepts such as development, heredity, or regeneration, which are essentially biological.
The effect on his theory of development of such acceptance of mechanistic principles is very obvious. This is why he clings to the conception of germ-plasm as the ground of heredity and the causal agent of development. It is the nearest he can get to a configurational explanation of these phenomena. If the development of a particular kind of organism starts out from a particular material organization,
1 In De la Metbode dans les Sciences , and edit., Paris, 1910, p. 25.
MISUSE OF ABSTRACTION 161
the germ-plasm of the species, and if this germ-plasm is reproduced or continued unchanged to form the startingpoint of the next generation, then hereditary resemblance is satisfactorily explained on mechanistic principles. If development is due to the formative influence of the germ-plasm it is understandable that it follows the same general course from one generation to another in any normal environment. 1 Actually the germ-plasm is a completely abstract and hypothetical substance, and the explanation offered of development and heredity is equally abstract and formal. The concept of germ-plasm is derivative from a mechanistic philosophy and has no other justification. It is, as we have seen, arrived at in the particulate theories by way of biological atomism, but the fundamental background is the mechanistic philosophy.
With the mechanistic scheme the concept of potentiality is incompatible; it is for this reason that Weismann, for example, disregards the plain evidence of the senses that development is an actual new creation of form and an actualization of functional potentialities, and postulates instead an initial material complexity which simplifies itself in the course of development, inverting thus the true sequence of events. The mechanistic philosophy can take no account of the historical aspect of vital activity; it must assume that past history is summed up in the present constitution of the germ-plasm. 2
It is for this reason that mechanists like Semon postulate in the germ-plasm material traces or cngrams to represent in the present the past experience of the race. Engrams materialize the past, just as representative particles or determinants materialize the future (cf. Delage, p. 81 above).
As we shall see in more detail in the next chapter, an organismal theory, which makes no abstraction of the psychical aspect but regards the living thing as a ‘psycho 1 Logically, he should include environment as part of the primary configuration.
2 Cf. McClung: *. . . the germ cells are old and strongly established in their characters — the resultants of millions of years of repeated experiences which are registered in their structure*. 1924, p. 650.
1 62 RETROSPECT
physical’ or organic whole, manifesting a distinctive kind of activity which is different from that of inorganic units (and may therefore properly be called ‘organic’), offers a better line of approach to the problems of development and heredity, and holds out hope of a less abstract and more adequate treatment of these problems. It will enable us to deal more satisfactorily with the all-important temporal relations of organic activity, to regard the living thing as a real continuing unity, extending through time. For the abstract concept of germ-plasm it will substitute the more concrete conception of the ovum as itself an organism — the future organism in its simplest state — having the elementary powers and functions of an organism, and the potentiality of developing the physical and psychical capabilities of the adult.