The Ghost in the Machine

  Docility and Determination

  It takes fifty-six generations of cells to produce a human being out of a single, fertilised egg-cell. This is done in a series of steps, each of which involves (a) the multiplication of cells by division, and the subsequent growth of the daughter-cells; (b) the structural and functional specialisation of cells (differentiation); and (c) the shaping of the organism (morphogenesis). Needless to say, all three are complementary aspects of a unitary process.

  Morphogenesis proceeds in an unmistakably hierarchic fashion. The development of the embryo from a shapeless blob to a roughed-in form, and through successive stages of increasing articulation, follows the familiar pattern described in previous chapters; I have mentioned the analogies with the sculptor, who carves a figure out of a block of wood, and with the spelling out of an amorphous idea into articulate phonemes. The step-by-step differentiation of cell-groups up to their ultimate specialisation presents the same hierarchically arborising picture:

  (after Clayton [2]). Diagram of some of the pathways open to early ectoderm in the amphibian embryo. Three only of the many inductive relationships are indicated by arrows.

  The diagram schematises some of the developmental possibilities of the ectoderm in the amphibian embryo. (The ectoderm is the outermost of the three layers of cell-populations into which the embryo differentiates at an early stage; the other two are mesoderm and endoderm.) The arrows on the left side of the diagram indicate the action of certain adjacent tissues ('inducers') which, when brought into contact with the ectoderm, act as chemical triggers on it. Those regions of the ectoderm which are in direct contact with the inducer-tissue will differentiate by stages into the animal's nervous system, including brain and eye-cups. Other regions of the ectoderm will, owing to their different surroundings, specialise in other ways. If a cell-population develops into 'skin', it may further specialise into sweat glands, horny layers, and so on. At each step biochemical triggers and feedbacks determine which of the alternative developmental pathways among several possibles a group of cells will actually follow.

  Thus, when the eye-cups (the future retina), which grow out of the brain at the end of two stalks (the future optic nerves), make physical contact with the surface, the skin over the contact area folds into the concave cups and differentiates into transparent lenses (see arrows on the fight of the diagram). The eye-cup induces the skin to form a lens, and the lens in its turn induces adjacent tissues to form a transparent horny membrane, the cornea. Moreover, if an eye-cup is transplanted under the skin on the belly of a frog embryo, the skin over it will obligingly differentiate into a lens. We may regard this obligingness or 'docility' of embryonic tissue, its readiness to differentiate into the kind of organ best suited to the tissue's position in the growing organism, as a manifestation of the integrative tendency, of the part's subordination to the interests of the whole.

  But 'docility' is again only one side of the picture; the other is 'determination'. Both are technical terms. 'Docility' means the multipotential capacity of embryonic tissue to follow this or that branch of the developmental hierarchy according to circumstances. But along each branch there is a point of no return, where the next developmental stage of the tissue is 'determined' in an irreversible way. If, at the earliest, so-called 'cleavage stage' of its development, a frog-embryo is split into two, each half will develop into a complete frog, not, as it normally would, into a half frog. At this stage each cell, though it is a part of the embryo, has retained the genetic potential to grow, if need be, into a whole frog -- it is a true, Janus-faced holon. But with each step of development along the branching tree, the successive cell-generations become more specialised, and the developmental 'choices' before a given cell-tissue -- its genetic potential -- become more and more restricted. Thus a piece of the ectoderm may still have the potentiality to develop into a cornea or skin-gland, but not into a liver or lung. Specialisation, here as in other fields, leads to a decrease in flexibility. One might compare the process with the series of curricular choices which face the student, from the first broad alternative between Science and the Humanities, to the final 'determination' which turns him into a marine zoologist specialising in echinoderms. At each point of decision, where the pathways diverge, some minor hazard or incident may act as a trigger which 'induces' him to make this or that alternative choice. After a while, each decision becomes to a large extent irreversible. Once he has become a zoologist, there are still numerous pathways of specialisation open to him; but he can hardly retrace his steps and become a barrister or a theoretical physicist. Here, too, the 'one-step rule' of hierarchies applies.

  Once the future of a tissue's development is decided, it can behave in a strikingly 'determined' way. At the gastrular stage, when the embryo still looks like a partly infolded sac, it is nevertheless possible already to tell which organs each region will produce. If at this early stage a piece of tissue from an amphibian embryo, which would normally give rise to an eye, is transplanted onto the tail end of another, older embryo, it will become, not an eye, but a kidney-duct or some other organ characteristic of that region. But at a later stage in the embryo's growth, this docility of the presumptive eye-region is lost, and no matter to what location it is transplanted, it will develop into an eye -- even on the host's thigh or belly. When a cell-group has reached this stage, it is called a morphogenetic field, organ-primordium, or bud, as the case may be. Not only the future eye, but a limb-bud too, transplanted to a different position (on the same, or on another embryo), will form a complete organ; even a heart may be formed on the host's flank. This 'ruthless' determination of morphogenetic fields to assert their individuality reflects, in our terminology, the self-assertive principle in development.

  Each morphogenetic field or organ primordium displays the holistic character of an automous unit, a self-regulating holon. If half of the field's tissue is cut away, the remainder will form not half an organ but a complete organ. If, at a certain stage of its development, the eye-cup is split into several isolated parts, each fragment will form a smaller, but normal, eye; and even the artificially scrambled and filtered cells of a tissue will, as we have seen (page 69), re-form again.

  These autonomous, self-regulating properties of holons within the growing embryo are a vital safeguard; they ensure that whatever accidental hazards arise during development, the end-product will be according to norm. In view of the millions and millions of cells which divide, differentiate, and move about in the constantly changing environment of fluids and neighbouring tissues -- Waddington called it 'the epigenetic landscape' -- it must be assumed that no two embryos, not even identical twins, are formed in exactly the same way. The self-regulating mechanisms which correct deviations from the norm and guarantee, so to speak, the end-result, have been compared to the homeostatic feedback devices in the adult organism -- so biologists speak of 'developmental homeostasis'. The future individual is potentially predetermined in the chromosomes of the fertilised egg; but to translate this blueprint into the finished product, billions of specialised cells have to be fabricated and moulded into an integrated structure. The mind boggles at the idea that the genes of that one fertilised egg should contain built-in provisions for each and every particular contingency which every single one of its fifty-six generations of daughter cells might encounter in the process. However, the problem becomes a little less baffling if we replace the concept of the 'genetic blueprint', which implies a plan to be rigidly copied, by the concept of a genetic canon of rules which are fixed, but leave room for alternative choices, i.e., flexible strategies guided by feedbacks and pointers from the environment. But how can this formula be applied to the development of the embryo?

  The Genetic Keyboard

  The cells of an embryo, all of identical origin, differentiate into such diverse products as muscle cells, several varieties of blood cells, a great variety of nerve cells, and so on, in spite of the fact that each of them carries the same set of hereditary instructions in its chr
omosomes. The activities of the cell, whether in embryo or adult, are controlled by the genes located in the chromosomes.* But since we have evidence that all cells in the body, whatever their function, contain the same complete set of chromosomes, how can a nerve cell and a kidney cell fulfil such different tasks, if they are governed by the same set of laws?

  * To complicate matters, there also exist cytoplasmic carriers of heredity, but for our present purpose these can be left out of account.

  A generation ago the answer to this question seemed to be simple. I shall put it into a somewhat frivolous analogy. Let the chromosomes be represented by the keyboard of a grand piano -- a very grand piano with thousands of keys. Then each key will be a gene. Every cell in the body carries a microscopic but complete keyboard in its nucleus. But each specialised cell is only permitted to sound one chord, according to its speciality -- the rest of its genetic keyboard has been inactivated by scotch tape. The fertilised egg, and the first few generations of its daughter cells, had the complete keyboard at their disposal. But successive generations have, at each 'point of no return', larger and larger areas of it covered by scotch tape. In the end, a muscle cell can only do one thing: contract -- strike a single chord.

  The scotch tape is known in the language of genetics as the 'repressor'. The agent which strikes the key and activates the gene is an 'inducer'. A mutated gene is a key which has gone out of tune. When quite a lot of keys have gone quite a lot out of tune, the result, we were asked to believe, was a much improved, wonderful new melody -- a reptile transformed into a bird, or a monkey into a man. It seems that at some point the theory must have gone wrong.

  The point where it went wrong was the atomistic concept of the gene. At the time when genetics got into its stride, atomism was in full bloom: reflexes were atoms of behaviour, and genes were atomic units of heredity. One gene was responsible for the colour of the eyes, a second for smooth or kinky hair, a third for causing bleeding sickness; and the organism was regarded as a collection of these mutually independent unit-characters -- a mosaic of elementary bits, put together in the manner of Mekhos' watches. But by the middle of our century, the rigidly atomistic concepts of Mendelian genetics had become considerably softened up. It was realised that a single gene may affect a wide rage of different characteristics (pleiotropy); and vice versa, that a great number of genes may interact to produce a single characteristic (polygeny). Some trivial characters -- like the colour of the eyes -- may depend on a single gene, but polygeny is the rule, and the basic features of the orgasm depend on the totality of genes -- the gene-complex or 'genome' as a whole.

  In the early days of genetics, a gene could be 'dominant' or 'recessive', and that was about all there was to it; but gradually more and more terms had to be added to the vocabulary: repressors, apo-repressors, co-repressors, inducers, modifier genes, switch genes, operator genes which activate other genes, and even genes which regulate the rate of mutations in genes. Thus the action of the gene-complex was originally conceived as the unfolding of a simple linear sequence like that on a tape-recorder or the Behaviourist's conditioned-reflex chain; whereas it is now gradually becoming apparent that the genetic controls operate as a self-regulating micro-hierarchy, equipped with feedback devices which guide their flexible strategies.* This not only protects the growing embryo against the hazards of ontogeny; it would also protect it against the evolutionary hazards of phylogeny, or random mutations in its own hereditary materials -- the blind antics of the monkey at the typewriter.

  * Significantly, Waddington calls his important book on theoretical biology The Strategy of the Genes (1957).

  At the time of writing, this kind of suggestion still meets with scepticism among the hard core of orthodox geneticists -- mainly, perhaps, because its acceptance must lead to a decisive shift of emphasis in our conception of the evolutionary process, as we shall see in the next chapter. But atomism, at least, is on its way out; it is encouraging to read, for instance, a passage like the following, quoted from a recent textbook for college students:

  All genes in the total inherited message tend to act together as an integrated whole in the control of [embryonic] development. . . . It is easy to fall into the habit of thinking that an organism has a set number of characteristics with one gene controlling each character. This is quite incorrect. The experimental evidence indicates clearly that genes never work altogether separately. Organisms are not patchworks with one gene controlling each of the patches. They are integrated wholes, whose development is controlled by the entire set of genes acting co-operatively. [3]

  Since differentiation and morphogenesis proceed in hierarchic steps, this co-operative activity of the gene-complex must also proceed in a hierarchic order. The gene-complex is enclosed in the nucleus of the cell. The nucleus is surrounded by the cell-body. The cell-body is enclosed by a membrane, which is surrounded by body fluids and by other cells, forming a tissue; this, in turn, is in contact with other tissues. In other words, the gene complex operates in a hierarchy of environments (page 102).

  Different types of cells (brain cells, muscle cells, etc.) differ from each other in the structure and chemistry of their cell-bodies. The differences are due to the interaction between gene-complex, cell-body and the cell's environment. In each growing and differentiating tissue a different portion of the total gene-complex is active -- only that branch of the gene-hierarchy which is concerned with the functions assigned to the tissue in question; the remainder of the genes is 'switched off'. And if we inquire into the nature of the agency which switches genes on and off, we find once more the familiar devices of triggers and feedbacks. The 'triggers' are the chemical 'inducers', 'organisers', 'operators' and 'repressors', etc., already mentioned. Needless to say, the way they work is only very imperfectly understood, and the proliferation of new terms is sometimes just a convenient method to mask our ignorance of details. But we know at least the broad principles involved. It is a process running in circles -- in circles which get narrower, like the coils of a spiral, as the cell becomes more and more specialised. The genes control the activities of the cell by relatively simple coded instructions which are spelt out in the complex operations of the cell-body. But the activities of the genes are in turn guided by feedbacks from the cell-body, which is exposed to the hierarchy of environments. This contains, apart from chemical triggers, a number of other factors in the 'epigenetic landscape' which are relevant to the cell's future, and about which the genes must be informed. To use a term proposed by James Bonner [4], the cell must be able to 'test' its neighbouts 'for strangeness or similarity, and in many other ways'. By feeding back information on the lie of the land to the gene-complex, the cytoplasm thus co-determines which genes should be active and which should be temporarily or permanently switched off.

  Thus ultimately a cell's fate depends on its position in the growing embryo -- its exact location in the epigenetic landscape. Cells which are members of the same morphogenetic field (for instance, a future arm) must have the same genetic orchestration and behave like parts of a coherent unit; and their further specialisation into 'solo players' (individual fingers) will again depend on their position within the field. Each organ-bud is a Janus-faced holon: relative to its earlier stages of development its destiny as a whole is irrevocably determined; but relative to the future, its parts are still 'docile' and will differentiate along the developmental pathway best suited to their local environments. 'Determination' and 'docility', self-assertive and integrative potential, are two sides of one medal (and so are, in the terminology of a hoary controversy among biologists, 'regulative' and 'mosaic' development).

  In the types of hierarchies discussed before, the time factor played a relatively subordinate part. In the developmental hierarchy, the apex is the fertilised egg, the axis of the branching tree is the progress of time, and the levels of the hierarchy are successive stages of development. The structure of the growing embryo at any given moment is a cross-section at right angles to the time axi
s, and the two faces of Janus are turned towards the past and the future.

  Summary

  The purpose of this chapter was not to give a description of embryonic development, but to point out the basic principles which this development has in common with other forms of hierarchic processes discussed in previous chapters. J. Needham once coined a phrase about 'the striving of the blastula to grow into a chicken'. One might call the ensemble of devices which make it succeed the organism's 'prenatal skills'. To quote James Bonner again: 'We know that nature, like man, accomplishes complex tasks by breaking them up into many simple sub-tasks.' [5] Development, maturation, learning and acting are continuous processes, and we must expect therefore that pre-natal and post-natal skills are governed by the same general principles.

  Some of these principles*, which we found reflected in embryonic development, were: the hierarchically branching order of differentiation and morphogenesis; the 'dissectibility' of that order into self-regulating holons at various levels (stages); their Janus character (autonomy versus dependence, determination versus docility); their fixed genetic canons and adaptable strategies guided by feedbacks from the hierarchy of environments; the action of triggers (inducers, etc.), which release pre-set mechanisms, and of scanners ('tests') which process information; the decrease of flexibility with increasing differentiation and specialisation. Lastly, we found earlier on that the canon of fixed rules which governs a skill is a 'hidden persuader', which operates automatically or instinctively. Mutatis mutandis, we may say that an analogous relation prevails between the genetic code of ancient origin and the 'pre-natal skills' of the growing embryo.