212 Within the Gates of Science and Beyond

In summary, this has been a very sketchy attempt to show that one
can seek and find the roots of art and its philosophy in nature. Man
with his powers of perception and aesthetic appreciation is one of
nature's products, just as much as are the patterns of organic forms
which he perceives in nature around him-the take-off points for his
One Plus One Does Not Equal Two *
flights of artistic imagination; hence they are both of the same kind of
origin and order. If in the limited space of a brief essay it has been
impossible to present more than the seed of the idea, perhaps some In school we learned that one apple plus one apple makes two apples.
future effort by one more competent to do so may bring it to One apple and one pear is just one apple and one pear. If we choose to
germination. ignore "appleness" and "pearness," being concerned only with weights
and numbers, as, for instance, in counting and weighing parcels for
postal shipment, apples and pears would, of course, be reduced to just
so many items, to be tallied by sheer summation. In equating such
items, we gain as well as lose. We gain an easy way of measurement, but
lose what nowadays would be called "information content." Apples
and pears do not become alike; we simply discount their differences for
a particular purpose.
But can we ever retrieve information about distinctive features once
we have tossed it out? If not, can science, as man's striving for as
complete and rational a picture of the universe as is obtainable to him
by observation, experiment, and logic, stoop to trading loss of
information content for the simplicity, convenience and, yes, true
elegance, of blotting out distinctiveness based on disparity; for instance,
between pears and apples? In fact, in nature, even two apples cannot be
equated, if one lies rotting on the ground while the other, still growing,
hangs on the tree.
In short, all algebra applied to nature implies abstraction. Sheer
adding up always leaves out some relevant information. Whether such
omission is passable depends on our purpose, and that, in turn, depends
entirely upon our special interest. And since interest, by definition,
connotes biased self-limitation, the information thus gained remains
incomplete, short of the comprehensiveness to which science in its
professed universality aspires. So, how sure can we be that sheer
analysis alone-the physical or mental dissolution of a complex into a

*Reprinted from THE NEUROSCIENCES: A Study Program (Ed.: Gardner C.

Quarton, Theodore Melnechuk, Francis O. Schmitt) Rockefeller University Press,
1967. pp. S01-S21.-Also in P.A. Weiss, LIFE, ORDER, AND UNDER-
STANDING. Special Supplement to THE GRADUATE JOURNAL, vol. S.
Austin, Texas, 1970.
213
214 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 215

shambles of measurable but disconnected units-does not irretrievably adding up and piecing together the microinformations about the
destroy highly relevant data about nature? Is Phoenix, rising from its smallest sample units. Never mind that physics had to give up that claim
ashes, a true image of nature or just a myth? gradually as Boltzmann's thermodynamics, Planck's quantum theory,
In our day, the answers to such questions have become a matter of and Heisenberg'S uncertainty principle came on the scene. The life
faith. The success story of learning more and more about less and less, sciences have failed to follow suit and break out of the strait jacket of a
which in the present context means about ever smaller fragments of doctrine for which their own subject matter furnished the most telling
nature, has grooved our faith in nature as an assembly plant of disproof. They might have come around more readily, though, if they
microevents. No doubt faith in the omnipotence of analytical decom- had realized that systems with aspects of wholeness are by no means
position has opened the mainsprings for the stream of scientific confined to living nature, but are of universal occurrence. In fact, their
progress. What we are apt to overlook in our enthusiasm is that there very universality should clear them of the stigma of vitalism.
are other sources which could powerfully augment that stream were Let me take a further step toward de stigmatization by pointing to a
they not left to dissipate and dry up for doctrinal reasons. Doctrine veiled source of confusion that seems to have confounded past dealings
has barred them from joining the mainstream by artificially erected with the problem-the failure to distinguish between a natural
walls, by conceptual injunctions against admixtures from sources phenomenon as such and the symbols of language we have to use in
suspected as contaminated because they failed to pass the orthodox test order to describe it. A phenomenon to which we ascribe wholeness is
of purity, namely, that one plus one must be made to equal two. certainly not more in algebraic terms than the sum of elementary
The unorthodox dissenters usually phrased their argument in the phenomena composing it. It just is different. The difference is that
age-old adage that "the whole is more than the sum of its parts." Look between matter and structure. If there is a "more" involved, it lies in
at this phrasing and you will discover the root of the distrust, and the terms of our description. It is we who, as describers, feel compelled
indeed, outright rejection, of the valid principle behind it. What did to add extra terms of information for the sake of making the
they mean by stating that "an organism is more than the sum of its cells description of the integral phenomenon complete and pertinent.
and humors"; that "a cell is more than its content of molecules"; that This neutral and philosophically noncommittal characterization of
"brain function is more than the aggregate of activities of its the problem tries to allay, or if you prefer, circumvent, the present
constituent neurons"; and so on? As the term "more" unquestionably warfare of dogmas. It should soothe the apprehensions of those who
connotes some tangible addition, an algebraic plus, one naturally had to have built faith in absolute reductionism as bulwark against onslaughts
ask: "More of what? Dimensions, mass, electric charges?" Surely none on their sense of intellectual security, and it should assure those others
of those. Then what? Perhaps something unfathomable, weightless, who felt disenfranchised because of their holist faith, their day in court.
chargeless, nonmaterial? All sorts of agents have indeed been invoked in There is a current fad to present the subject matter of the life sciences
that capacity-entelechy, elan vital, formative drive, vital principle-all in terms of a dogmatic schism-an antithesis of "molecular" and
idle words, unpalatable to most scientists for being just fancy names for "organismic" biology, professing a reductionist and a holistic philos-
an unknown x. ophy, respectively. The former is respected for its "rugged naturalism,"
Unfortunately, in their aversion to the supernatural, the scientific the latter suspected of flirting romantically with the supernatural. What
purists poured out the baby with the intellectually soiled bath water by ,- I shall try to show is that exclusive commitment to either thesis is
repudiating the very aspects of wholeness in nature that had conjured unnatural. The molecular and the organismic are but two different
up those cover terms for ignorance. And as a prophylactic against their vantage points from which to look at living systems, neither of them
resurgence, they fostered a militantly doctrinaire "reductionism," granting a monopoly to insight. They are complementary and co-equal.
which axiomatically prescribed that all the relevant macroinformation To document this proposition is the main object of the following
about nature must, and eventually will, be derived completely from' discourse; I hope that it will serve as an object lesson.
216 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 217

Specifically, these are the points I aim to prove: (1) that as our
-•.,•.,•.,•.•
·•••••••
-•.•.••.••.• .•.••
brain scans features of the universe we shift range and focus back and
forth between telescopic and microscopic vision, as it were; (2) that as
we move downward on this scale, we mostly gain precision and lose
II • • • • • •
perspective; (3) that as we move upward, new and relevant features,
formerly unrecognizable and unsuspected, come into view; (4) that this
emerging novelty pertains to macrosamples of nature-that is, that it ••
•••••••••
reflects properties of collectives-of groups, assemblies, systems, and
populations, composed of microsamples; and (5) that the, required
additional terms to characterize such collectives must come from
rigorous scientific procedure rather than from anthropomorphic trans-
locutions and allegorical allusions to mythology.
·•.....
• •••••
•••••
• ••••
II • • • •
~
I

•-....
And now I turn from these somewhat pedantic generalities, which

••••
to some presumably will seem commonplace, to practical examples by • • • • • I

which the validity of those five points can be tested. A brief glossary of
our terms of reference may serve as introduction. Let us ask first: Of
..... _I
• II • • • j
what do we deprive a system when we dismember it and isolate its
component parts, whether bodily or just in our mind? Plainly, of the
interrelations that had existed among the parts while they were still
united. So, in trying to reconstruct the system from the fragments,
whether bodily like Humpty-Dumpty, or symbolically in our imagina-
y~~.:-
Y~jfJl~~J
...-:j
tion, we must make up for the deprivation by adding a proper term that
Figure 1
specifies the lost relations. This may simply amount to adding vectors
to algebraic terms. The requirements for added specifications will vary
with the different degrees of order emerging from the union (or
reunion) of elements combined in groups. The simplest case involves
only a loose and widely variable relation, such as "togetherness"; it Figure 1 shows a meaningless array of dots in inert coexistence,
displays novelty, but little order. If, besides novelty, the collective with nothing recognizably in common besides the paper they were
shows regularities of pattern which recur with a high degree of printed on. Yet, from a greater distance or, what is the same, at lower
invariance, we confer upon it the designation "organized." magnification (inset), we recognize them as the component bits of
But here again we would do well to make a further distinction information about a continuous, well-structured image. That image, of
between true and merely simulated organization. We must distinguish course, is dead; the dots of printer's ink composing it are physically as
between the genuine order, such as emerges within a group by virtue ~f unrelated as fly specks. What gives the picture its meaningful
its intrinsic dynamics, and a mere semblance of order, such as an integration, are we, the viewers, with our eyes and brain. The dots do
aggregate of unrelated units acquires by imposition or imputation from not "add up." We add to them. From this we learn that discontinuous
without. Examples of the latter are puppets, or the proverbial camel and discrete elements can give us the illusion of continuity, but that the
our fantasy projects into a cloud, or, in fact, any effigy of a natural mere aspect of continuity alone is no test of inner coherence. Let me
system, as in the following instance. pursue this further.
218 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 219

Figure 2 is a picture of the spiral galaxy, Andromeda. Now, if you

ask what Andromeda consists of, a census taker would reply,
"So-and-so many stars"; a chemist might come forth with spectro-
scopically discernible distinctions. But would either of these answers
add up to a definition of a spiral pattern? Certainly not. Nor would a
scientist be happy if the additional feature were taken care of by
invoking the magic act of "a spiralizer."

Figure 3

Closer to earth, Figure 3 shows photographs taken by the first

weather satellite, Tyros. Note the cyclonic cloud pattern. But what are
clouds in analytic view? Droplets of water. Now, could knowing all
there is to be known about H 2 0 ever add up to a picture of this
configuration? Of course not: the winds that have shaped it remain
invisible. In Christina Rossetti's words,

"Who has seen the wind? Neither you nor I.

But when the trees bow down their heads,
the wind is passing by."

So, here we meet the first caveat against willful isolation of an

Figure 2 object from its natural context.
220 Within the Gates of Science and Beyond
One Plus One Does Not Equal Two 221

channel further flow, as in this instance (Figure 5) of a spiral spout

carved by the twisted course of water that drained from a glacier
bottom in the glacial age. Once formed, the spiral structure becomes
self-perpetuating, gaining in polish and perfection by more erosion.
In passing, let me point out that by this dual action, a whorl can
serve as a general model of how dynamic patterns tend not only to
preserve, but in further consequence to accentuate, their self-
engendered structures. I wish to stress this because of the obvious
bearing the general two-step principle has on our understanding of brain
mechanisms: of grooving, habit formation, facilitation, and learning
theory. If taken seriously and followed up, it might dispose of the
necessity of placing rigid fixation and plasticity of neural functions-
instincts versus memory-into sharply divided categories, each run on a
different principle.
Returning to our main line, we have recognized the spiral wall of
Figure 4 the glacial mill as the dead effigy of the unitary dynamic sweep pattern

More spirals next, in Figure 4: the neurofibrils in the large motor

cells innervating the electric organ of the torpedo fish. To say that these
fibrils are made of protein will neither describe nor explain their spiral
course. Indeed, from what we now know about their formation, the
picture is just a photographic still, a momentary sample of a
continuously unwinding record of motion in that cell. These spirals are
but the residual traces of a moving stream of substance, the pattern of
which must be sought in regularities of the underlying dynamics.
Here lies a general and basic lesson. What we perceive as static form
is but the product, transitory or lasting, of formative processes. The
features of the product-for instance, its geometry-can provide us with
clues for the dynamics that underlie those processes. For instance, the
counterclockwise spiral spin of water running off a bathtub drain is but
an indicator of an asymmetry of forces resulting from the earth's
rotation. The order we perceive in structured form thus is not primary,
but the expression of the dynamic patterns that have engendered it. Yet
this is only a beginning. Through their results, dynamics modify the
setting for subsequent dynamics. Dynamically created forms, if
somehow consolidated, become molds for the course of further
activity. Free-flowing water grooves its bed until the bed begins to
Figure 5
222 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 223

that has created it. The spiral composition of an artist (Figure 6) is, one is faced with static geometric regularities of patterns, he ought to
similarly, the projection on dead canvas of some dynamic process, look beyond them-or, rather, behind them-for the rules in the play of
obeying mathematical terms for spirality, that has been going on in the forces that have shaped them. In thus raising the sights from statics to
artist's creative brain. Attempts to resolve this act to mere terms of dynamics, static interrelations become dynamic interactions, and in the
numerical plurality, whether of neurons or of intraneuronal molecules, case of self-sustaining systems with the conservative features of
would seem to me to be as futile as to derive the spirality of a spiral wholeness, simple interactions become interdependencies. States then
nebula from our knowl~dge of single isolated stars. appear as but cross-sections through trains of behavior along the
By now I have exposed three propositions. First, that collectives time-line, scalar values must be supplemented by vectorial interconnec-
tend to display novel features not discernible in their component units, tions, and vector systems of specifiable integral properties become
hence justly called "emergent"; second, that such features are indicative realities. Let us then keep in mind that this progression from elements
of the existence of significant relationships among the members of the
to groups objectively reflects the ascending scale of su~plem~ntal
collective, such relationships being severed by physical or mental statements we need for adequate description of correspondmg objects
separation of the members from each other; and, third, that whenever of our experience. I shall then present samples of such phenomena in
that oraer. By choosing them from various points along that scale, I
intend to blur the artificial dichotomy between modes of thought
centered either on elements or on continua, each to the exclusion or
invalidation of the other.
I shall use the example of form as master indicator of order. Its
simplest examples are plain aggregates of identical units stacking up
flank-to-flank or end-to-end, according to steric fitting, like key to lock,
and chemical conformances. The macromolecular units of the blood
pigment of a marine worm (Figure 7), each consisting of six subunits

Figure 6 Figure 7
224 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 225

ordering. For short lengths, intermolecular forces may serve as an

explanation, but ruler-straight arrays over great distances undoubtedly
must be referred to straightening effects from the environment; for
instance, stretch. Common direction may be imposed by a further
polarizing interaction with the environment. And so, in order to
describe the formation of, for instance, a connective-tissue fiber, we
must construct steps of ever more specifications.
The stacking of lipid molecules into lamellar systems of the
so-called smectic state (Figure 9) extends the same principle to two
dimensions. Here, too, the environment enters as an ordering factor in
that it offers to the molecules, as scaffolding for their own planar
self-array, a planar interface between two immiscible media. Lacking
such guidance, the molecules cluster into so-called micelles (Figure 10),
Figure 8 yet by no means as random conglomerations, but in orderly structural
patterns determined by their own collective interactions. For each of
them the others are part of its environment-a forward reference to our
conclusion that the notion of independent "elements" is, in itself, an
abstraction, for in reality elements are part and parcel of a single,
undivided continuum that embraces units and environment as one
integral entity.
Yet, clearly, structural group order in these past examples can still
be satisfactorily explained in essence by the microprecise automatic
assembly of individual units and subunits, united like an erector set by
their steric, chemical, and electrical properties. Such rigid compounding
processes can hardly serve as star witnesses in our suit for a divorce
Figure 9 from the one-plus-one-equals-two precept of thought. Therefore, let me
proceed to a series of further samples, representative of higher-order
around a hole, stack up in contact: molecules as "modules." The only systems. These are collectives of the following description: their
novelty by which the group differs from a mere sum of units is its features, "on the whole," show well-defined regularities of pattern,
predominantly planar array-a significant, yet low, degree of order. The recurring consistently from specimen to specimen in each given class;
stacking of virus particles (Figure 8), in its near-crystalline configura- but as one looks at smaller and smaller samples, similitude and
tion, falls in the same class, although the forces interlinking the units so regularity decline until, having descended to the elements, one can no
regularly are not equally obvious. longer find any hint of what the structure of the total complex might
A polymer, as I shall show later, is a linear chain of identical links be like. This is because the details of pattern are in each case unique, no
with couplings, end-to-end, like a railroad train. The linkage represents two microsamples being ever alike, even though the composite pattern
a first step of order in the assembly. The straightening of the chain of the whole, case after case, is of the same standard form. Contrary to
from random coils to rectilinearity requires an additional step of the preceding examples, the order of the whole can here no longer be
226
Within the Gates of Science and Beyond
One Plus One Does Not Equal Two
227

predicted from a simple upward projection of the elemental properties

of single units stacked up in module fashion. The following illustrations
will make this more specific.
Take, for instance, the bed of blood capillaries in a tissue (Figure
11). They branch and re-anastomose almost at random, yet the
resultant network offers an aspect of great over-all regularity. Descrip-
tively, the regularity is reflected in the near-constancy of distances
between the branches; dynamically, it reflects a growth pattern
elaborated by interactions of the component branches, both among one
another and with the cellulated matrix they pervade. In oversimplified
terms, the interactions involved are a type of competition. This, then, is
interaction no longer in contiguity, but at a distance. Each branch may
be viewed as surrounded by a shell of influences of graded strengths-
domains, which keep each other at a respectful standard distance. The
term "domain" is used in forward reference to subsequent comments
on field and gradient principles.

Figure 10
Figure 11
228 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 229

Figure 12 Figure 13

Figure 15 Figure 16

The growth pattern of a nerve cell from the cerebellum (Figure 15)
reveals the same rule: the same degree of unpredictability of the details
of ramification, yet at the ends great uniformity of distribution among
the terminal branchlets. While it is their environment-the matrix of the
brain-that offers to the advancing branches a warp and woof of easy
pathways, the decision of which of them will be utilized, and in what
force and microdistribution, is indeterminate, left to be decided at each
branching point by the actual local competition for the limited supply
of substance arriving from the common cell body. So, if the common
source may have had a "program" for the attainment of the highly
Figure 14
regular end result, the precise way of how to get there could certainly
A botanical counterpart to the capillary bed is the venation of a leaf not have been spelled out in it in great detail. For those ways are
(Figure 12). But let me at once dispel the notion that growth patterns different and unique for each of the millions of cells. The double
of this type are a preserve of organisms. The next picture (Figure 13), meaning of the word "design" comes to our mind: design as purposeful
for instance, shows the lightninglike pattern of an electrostatic planning at the start, ending in stereotyped design as accomplishment,
discharge from a point source. Ideally, it should, of course, be radial but but countless ways of execution leading from one to the other.
in reality, the unpredictable variations of conductance and resistance, Extending our examples upward, Figure 16 pictures a lace coral-a
resulting from the random inhomogeneities of the medium through limestone housing development of both great over-all regularity and
which it has to travel, establish spearheads for separate and competing individual uniqueness, built by thousands of separate little animals in
ionization tracks. Despite this capriciousness in detail, the total picture the colony in a concerted pattern of behavior. You may sense already
still emerges as one of systemic order. Growth patterns of snowflakes my own design. It is to reorient thinking from static form to formative
(Figure 14) also show infinite variation of detail within a high degree of behavior across all orders of magnitude. The range extends beyond the
constancy of the over-all form of the growing crystal.
 230 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 231

Figure 17

coral colony to human society and to what I take to be its design for
living, and indeed, survival: namely, to recognize that individual
freedom in the small is compatible with the existence of collective
order in the gross, which reconciles self-determination of the individual
with the much stricter frame of rules descriptive of his group.
Were it not for this principle of nature, were the development of
every part or branch allowed to pursue its own capricious course
without constraints, without a frame of integral interdependencies, we
could not have trees (Figure 17) that we could categorize distinctly by
their shapes as oaks or pines or poplars even though each specimen is
individually unique. Such standardized end form defies any logical
attempt to regard the product as just the blind outcome of a bunch, or
call it a sum, of microprecisely programmed cause-effect sequences of
linear chain reactions in the sense of a naive mechanical machine
concept.
The conclusion that countless constellations of convergent micro-
events may yield macroproducts of essentially the same standard
pattern makes it, by the same token, gratuitous to assume that similar
terminal patterns must have had similar mechanisms and histories in
common. The treelike pattern of the Colorado River delta (Figure 18)
Figure 18
 232 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 233

Figure 20

Order emerging in complexes of disconnected, discrete units taxes our

explanatory faculties far more severely. As introduction to the subject,
I chose Figure 20-three frames from a motion picture film of one and
the same cell in tissue culture at brief intervals. They show a group of
granules, each about one micron in size. These granules are separate
bodies, freely mobile, bouncing around in the soft cell plasma. But as
they change positions, they assume preferentially characteristic geo-
metric group configurations-mostly hexagonal, but intermittently an
Figure 19
occasional square. Now, since they are separated by appreciable
distances, we must infer that in their random buffeting by Brownian
will prove the point. Let us go on then and reverse the outgrowth motion, they ate transitionally stabilized-trapped, as it were-at
pattern of the tree and we obtain the picture of a river forming by equidistant equilibrium points in a field of forces established by their
junctions of tributaries from many sources; but also, similarly, the mutual interactions, like partners in a square dance or quadrille. If this
inverse "arborization" of cracks in crystals advancing from edge to sounds vague, consider that we can at least describe the various fleeting
interior in stepwise confluence (Figure 19, from top to bottom). configurations with relative precision, even though we do not know the
Let us now move up to the next step of complexity. Our past actual dynamics defining the grid.
examples have been relatively simple. They dealt with interactions There are other cases, however, in which the operation of group
among parts of systems which, after all, were still connected and dynamics has lent itself to more concrete definition. Since they
continuous in substance-blood vessels, leaf veins, trees, rivers, etc. demonstrate most cogently that going beyond the one-plus-one-equals-
 234 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 235

two rule does not mean giving up scientific discipline for the outer
space of supernaturality, I shall dwell on them somewhat more
extensively.
Let us take two bodies (Figure 21), as centers of emanations and
force fields extending radially into the environment, and let them move
toward each other from a great distance. Beyond a certain range, (0)
interactions between the two arc as negligible as the effect of
gravitational attraction by the moon is on our stance. Yet as we bring
like sign unlike sign
them closer (Figure 22) and as the overlap of their domains increases,
their joint effects depart increasingly from the result one would expect
from a sheer superposition and algebraic summation of their single
contributions. They mutually distort each other's sovereign patterns of
action.
Proton No.: 2 2

Figure 23

Figure 21

"".

Figure 24
Figure 22
spherical electron clouds, as shown in Figure 23 (bottom line). The
In atomic dimensions, for instance, this yields the redistribution of dumbbell-shaped interaction pattern seen in the figure at the left is
electrons between atoms (ligand fields), deforming the erstwhile typical of many cases. The next picture (Figure 24) shows the mapping
236 One Plus One Does Not Equal Two 237
Within the Gates of Science and Beyond

.~/ .•..•. "'\1

.... (,

Figure 26 Figure 27

foci, or what not, corporeally isolated from their context, has long been
a legitimate abstraction of immense tactical and practical utility in
science. Yet, he who forgets that it is basically an abstraction could as
well end up trying to extract the center of gravity from a body. One
Figure 25 recognizes a kinship between this latter trend and some of the old
notions, still not totally extinct, about brain centers as the "seats" of
specific functions.
Familiar and accepted as the preceding proposItions are for the
of electron distribution in a small organic molecule (specifically a
molecular realm, their equal validity for higher levels, through the
diaminochloropyramidine). Such a continuous field pattern emerges
cellular to the social, has rarely been pointed up, let along studied and
from the group interaction of the constituent atoms and atomic groups,
conclusively proven.
which formerly were envisaged and represented as discrete, neatly
Let me again start from an inorganic model, the so·called rings of
bounded entities. More complex molecules-for instance, the protein,
Liesegang (Figure 26). A drop of silver salt dropped on a gela.tin plate
myoglobin, shown in Figure 25 in the so·called Patterson projection of
that had been soaked in a chromate solution lays down, as It slowly
its subunit fields-yield maps of still mdre sophisticated collective
diffuses, periodic concentric rings of insoluble silver chromate. The
fingerprints. One is reminded of the contour maps of mountain ranges.
rhythmic character stems from some sort of threshold phenomenon,
Domains of particles are no more truly isolated than are mountain
formally comparable to the rhythmic response of nerve tissue to a
peaks.
constant stimulus. If we place two such diffusion centers sufficiently
My reason for showing these diagrams is that they express
far apart on a common plate (Figure 27), the total pattern still adds up,
symbolically that patterned processes in space and time form continua.
in the main, to one.plus.one. Yet, if we narrow the original distance
To single out and fence in mentally, in such continua, peaks, centers,
 238 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 239

in the configuration of the Great Dipper. Figure 29 shows the outcome

of an actual experiment. As one notes, the resulting pattern of silver
lines coincides with the pattern of point connections that led man to
give the stellar group its name. Does not the unequivocality of this
correspondence intimate that a similar dynamic interaction pattern in
man's brain had guided his interpretation? Stars do register on retina
and brain as single points, but may not the several neural processes
thereby actuated engender, on a higher brain level, dynamic inter-
actions that integrate an erstwhile mosaic of local dots into a unitary
spatial pattern?
At times, proposals for such physiological underpinnings of Gestalt
phenomena have been set forth, couched mostly in symbolic terms of
field concepts and related models. Experimental verification, however,
Figure 28
has remained scanty. Because cell types other than neurons have
furnished far more factual examples of how pattern-determining field
effects can arise, I shall turn to those. I shall present two major types of
patterning interactions among dispersed cells, first, in a liquid (thermal-
ly agitated) environment and second, in a firm cohesive matrix.
The earliest morphogenetic step in the egg of the seaweed Fucus is
the sprouting of a rootlet on one side. In isolated eggs this sprouting
takes place at quite a random spot. When several eggs are combined in
groups, however, sprouting is patterned. As Whitaker has shown, pairs
of eggs in close proximity sprout at the sides facing each other (Figure
30). However, huddles of more than two sprout toward the geometric
center of the group (Figure 30). Thus "mutual attraction," which might
still have been conceivable within pairs, is clearly ruled out as
explanation for groups of more than two. How then can one account
for the phenomenon? Quite simply. When Whitaker placed undivided
eggs into a pH gradient (Figure 31), the rootlets sprouted at the more
acid side, perhaps because the egg membrane was weakened on that side
and sprang a leak through which the rootlet could hatch. Now, eggs
may be assumed to secrete their own acid as a metabolic product. This
Figure 29 acid cannot diffuse from the confined space within a cluster as rapidly
as it can from the outer shore. As a result, an inner-to-outer
between them, their mutual interference becomes conspicuous (Figure concentration gradient of acid will develop and polarize the members of
28): the twin set of circles becomes distorted, the more peripheral ones
the group toward their common center. The additional step that the
merge to single dumbbell-shaped contour lines, and the outermost and
experimenter had to take to localize root formation the eggs perform
farthest advanced form a smooth enclosure of both. The two domains
among themselves in concert, thus adding what makes two different
have fused. Carrying on the tests, I then placed seven drops on a plate
from one-plus-one, and even three different from the new two-plus-one.
240 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 241

,---

• • •.
• ," •••
• • A
~TROL HIGdJ
••
'" .'... ~
• •
....
Figure 30 Figure 31

Figure 33
: ... ~ ..~ .....
; 'tI0

Figure 32
<~~~
.......
_. .0
. inner medium and its outer environment apparently reversed the
gradient, and the dendrites consequently grew outward. The actual
agents involved here have not been determined. For other tissue cells,
however, we could prove (Figure 33) that bipolar cells in culture,
acidified at one end, withdraw that process, thus becoming unipolar.
This fairly reproduces what two cells exuding acid will do to each other
as soon as they come close enough. of course, exudates other than
In clusters of explanted embryonic nerve cells (Figure 32), acids could have the same polarizing group effect.
Stefanelli observed a similar convergent growth of dendrites toward the Group patterns among cells in semisolid media arise differently.
common center, but only if they were not near the outer edge of the Most tissue cells, as well as nerve fibers, need the support of solid
drop of culture medium; for near that border, competition between the structures-fibrin or collagen fibers, for instance, along which they
242 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 243

Figure 34

move and grow, like plants along a trellis. The diagram in Figure 34
summarizes the gist of four decades of experiments on this principle of
"contact guidance." An untreated protein coagulum, e.g., fibrin in a
blood clot, is a random tangle of fibers (top of figure). As illustrated in
the lower part, stretch orients the mesh in the di~ection of the lines of
stress. Depending for guidance, like blind men, on the fibrous tracks,
the cells then trace the underlying structure. Cell group patterns thus
have their precursors in the fibrous matrix in which they are enmeshed.
Therefore, if cells could do to the matrix what the experimenter does in
applying stretch, they could evidently manage to set up their own
physical interconnections and group patterns. And indeed, they can do
this. Here is how.
A cohesive fibrous network is under internal tension. Any local
disruption of the net makes the surrounding meshes retract to form a
ring around the hole, as any lady knows from holes in her stockings.
Some spiders (Figure 35) use this as a trick to build a strong-walled
nest. Now, cells can achieve the same result, where needed, by local
liquefaction through proteolysis of their matrix, or just by local
expansion (Figure 36 top). Fibrous and muscular coats around hollow
organs owe their circumferential orientation to this effect. More
pertinent for us here is the bottom diagram, which illustrates local
shrinkage. The meshes are gathered purse string fashion, assuming a
radial orientation with focus on the shrinking center. Now, some cells, Figure 36
244 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 245

Figure 37

especially proliferating ones, make their surrounding matrix shrink in

just this fashion by the release of chemicals that make the meshes Figure 38
squeeze out water, like setting jelly. If cells subsequently emigrate from
such a center, they naturally follow their self-created radial routes.
Now, if there are two such cell clusters in a common matrix, will
the resulting growth pattern turn out two overlapping stars? Of course
not. Let me recall the diagram with which I introduced the two-center
theorem earlier in the article (Figures 21 and 22). Two local centers of
contraction in the net decidedly do not add up in their effects, as can
be demonstrated readily by gathering a taut piece of mesh fabric at two
points (Figure 37): the meshes between the two centers are distorted
lengthwise into a straight course along the connecting line. In the same
way, two clumps of growing cells force fibers in their common colloidal
medium to assume a straightline orientation along the shortest distance
between them. Three centers generate a structural triangle in their
matrix (Figure 38), which then serves as roadway for emigrating cells.
Thus, three scalar and erstwhile unrelated local chemical activities
become upgraded through vectorial interaction into a well-defined
space pattern emerging de novo. Figure 39 shows an actual case of such
an automatically established triangular interconnection among three
embryonic spinal ganglia in vitro. Figure 39
248 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 249

pragmatic, conceptual artifact has serious limitations, and once we

reach the limits of its applicability we must relax our historical
commitment to exclusive legal recognition of conclusions arrived at
analytically and must concede equal explanatory status to collective
statements of fact in their own right. Modern physics has done it
implicitly in adopting thermo-dynamics, quantum theory, relativity,
and statistical mechanics, and the time has come when the life sciences
had better follow suit for their own good. They must learn to accept
recognized orderliness of the behavior of systems on its own terms. If
they can reduce it to analytical terms, well and good. If not, the reality
of nature still must be allowed to prevail and override pet micro-
mechanistic preconceptions based on predilections. Therefore, regard-
less of whether ordered behavior patterns of systems will yield to
obstinate analytical efforts at piecing them together through sheer
assembly of component pieces-whether of molecules joining to form
specific macromolecular systems, of cells to compose organs, or of
Figure 42 neuronal circuits to yield adaptive functions-the integral formulations
will retain their claim to reality and primacy. Both conceptually and
the most efficient structure of a modern growing bicentric city historically, reductionist description is a secondary and limited tactical
embodies a remarkably similar symmetric pattern of settlement along convenience. Its limits are to be determined empirically and not by
pressure and flow lines of communication. Population and market prejudice.
distribution between two actual towns, mapped to scale by Isard Let me briefly pursue this matter on the example of linear arrays,
(Figure 42) bears further witness to the operation of the two-center already mentioned briefly earlier. Collagen fibers (Figure 43) are
effect in the dynamics of human ecology. bundles of polymeric chains of protein molecules, linked head to tail
This mention of ecology brings me to my last set of examples. and flank to flank. Under certain conditions, homologous subgroups of
Ecology is group behavior in free interaction with other groups and neighboring chains line up in register, resulting in conspicuous
with environment. It epitomizes the lifting of subject matter from sheer cross-banding. Each molecule is characterized by a specific sequence of
catalogues of items to paramount concern with their typical differential amino acids along its backbone. The sequential order being commonly
distribution and, digging deeper, with the patterns of the underlying referred to as a code, it justly can be compared to the sequence of
dynamics. Behavior patterns are ecology'S instruments. To explain life, letters in a word. Register, then, signifies the sliding of identical letters
static cell anatomy must become molecular ecology, organisms be into alignment (Figure 44), and there is good rea~on to expect that the
comprehended through cell ecology, and societies through the dynam- molecular mechanism of zipping collagen fibril to fibril by site-
ics of human ecology. They all provide us with examples of rigorous matching will be revealed before long. This instance of higher group
scientific propositions that hold for groups but dissolve when efforts order then resolves itself simply into a case of assemblage by the
are made to reduce them to elemental properties. Here resolution orderly stacking of contiguous elements, not' unlike our initial
becomes sheer dissolution. examples. But what if similar ordered arrays arise in collectives of linear
Let us then be emphatic: True, scientific history has grooved our units without the benefit of mutual contact? How to explain, for
habit of explaining group behavior in terms of the interactive behavior instance, the corresponding group pattern of freely mobile, well-
of quasi-independent unit actors, whether molecules or men. Yet this separated, linear units in Figure 45, which shows an assembly of trout
250 Within the Oates of Science and Beyond One Plus One Does Not Equal Two 251

evenly spaced in parallel and register on the bottom of a stream?

Clearly, current flow lines and interindividual signals combine in
yielding this behavior pattern in grid form, but just how is wholly
obscure. How far, then, can we carry the notion of synthesis of such
ri,.~. !'t:' ~i.{~~ • ~ ; [i!' '!, 'for .,." ",- I i':.- .» group patterns through the free interaction of their unit elements?
I.. a
"':.';:: .
1, ,
, . . Figure 46 presents an aerial photograph of logging in the Columbia
~ l" I ~'> tt:. ~

• ..0"1 ,," ,'- • ':: • _~ ! ~ "t~, -, <;1 l.j . River. The logs are all aligned by current flow and shore lines. Their
deposit in register is man-made. In other words, the collective order is
Figure 43 strictly imposed by outside forces, which makes it irrelevant to our
present context. Not so the next example. Figure 47 is an electron
micrograph of rod-shaped tobacco mosaic virus. The units are clustered,
and within each cluster the component rodlets are again both in
alignment and in register. They look like match sticks; and indeed a
group of matches can by analogy serve as a model of the physical
mechanism through which such a simple step of order in self-assembling
groups can come about. If one scatters matches at random on the
surface of water in a dish and then agitates the surface by continuous
tapping on the container, the floating matches get into motion and as
they collide, they turn into positions of mutual alignment and
register-positions evidently satisfying an equilibrium (minimum surface
energy) requirement for that particular three-phase (water-wood-air)
system. Three stills from a motion picture film of such a model
Figure 44 experiment are reproduced in Figure 48, to show the progress in the
increase of order by "self-ordering."
Being an instance of ordered group behavior emerging visibly from
elementary interactions, the case surely is heartening to reductionist
faith. Unfortunately, the argument is open-ended, as can be readily
observed in motion pictures of a remarkable rod-shaped microbe,
Bacillus circulans. Loosely scattered bacilli start out by assembling in
physical arrays just like those matches (Figure 49), but once the group
has enlarged to a certain critical size, the whole mass adopts a totally
different course of behavior: it begins to circle around its own
geometric center as fulcrum and keeps on rotating for indefinite periods
of time (Figure 50), like a revolving disk, regardless of whether the
number of individuals are counted by the hundreds or hundred
thousands. There is no sign in the behavior of individuals before their
assembly that would have intimated the future rotatory performance of
Figure 45 the collective. In fact, although each spinning mass tends to keep its
252 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 253

Figure 48

Figure 46

Figure 49

Figure 47 Figure 50
 254 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 255

exercises is spectacular. Unfortunately, this running in circles finds a

counterpart in human affairs. In fact, in general, there is not too much
difference between the laws of human ecology and the ecology of cells
and molecules. Dependence on environment, self-sorting by segregation,
compounding in groups, recombining for symbiotic reciprocity-in
short, self-patterning of groups-occur among molecules and men alike.
Take, for instance, this aerial view of Coney Island on a Sunday
(Figure 52). Consider the people as molecules. The heavier border on
" top is the condensed belt of hydrophobic bodies adsorbed to the
".I ' water-beach interface. The dark clusters inside the mass clearly mark
I
domains of attractive forces, presumably emanating from sources of
nutrient and stimulant attractants. Their equidistant spacing indicates
mutual repulsion through forces of competion; and so forth. The
analogy is not at all facetious. It cuts deep into the heart of our topic,
for it exemplifies basic features of self-organizing systems. I could have
gone on, for instance, to relate how a random mixture of isolated single

Figure 51

sense of rotation once it has been started, whether clockwise or

counterclock~ise, .fragments that split off may change to spinning in
th~ reverse dlrectIOn. So evidently the individual units are totally
um~structed as to the direction in which they will be made to spin by
then assembled community. Here we are faced with true emergent
novelty .
. The circling motion of masses of army ants is equally instructive
(Flgure 51). They keep running round and round a central column
d.rawing in stragglers, leaving others behind, in an interminable mas~
clrcus movement. The infectious pattern of these insect dervish
Figure 52
 256 Within the Gates of Science and Beyond One Plus One Does Not Equal Two 257

cells (Figure 53), obtained by dissociating an already functioning

embryonic kidney, then scrambled, lumped and properly nourished,
can reconstitute itself into a remarkably well-organized miniature
kidney (Figure 54); or how similarly scrambled cells of embryonic
chick skin in tissue culture can grow into normal feathers (Figure 55);
all of this entirely by "do-it-yourself" methods.
Examples of "self-organization" of this kind are numerous. To label
them is easy, although gratuitous. To understand them is a long way
off. Exclusively reductionist tactics will never get us there, if they
persist in "going it alone"; nor, on the other hand, will sheer verbal Figure 53
soporifics. What the task calls for is, first of all, a job of thorough
conceptual overhauling and renovation. It requires that we drop
self-imposed blinders and admit to view the higher perspective of the
whole-not just its bogus literary versions, but its hard scientific core,
expressed in such phenomena of emergent collective order as I have
illustrated. The venal preoccupation with bits of the materials of nature
as such-with "what there is "-must give way to a broader concern with
the manner of their operation and use-with "how it all works." And,
in this shift of emphasis, one discovers that all the bits hang together; Figure 54
that they are all intermeshed in webs of subtle interactions forming
domains or subsystems within the over-all continuum of the universe.
To emphasize the "systems" character of the dynamics of living
entities I have, ever since 1923, couched their description in terms of
the concept of "fields." Lest this symbolic term again arouse alarm, let
me decontaminate it instantly by the following simple example (Figure
56). Let us take a circumscribed body, depending for its maintenance
on active exchange with its environment; for instance, an egg in the
ocean, a cell in a tissue, a human individual in society. Then let the unit
multiply into a few more units; they all continue to have a share in the
common interface of exchange and communication with the medium.
But let the number of units keep on increasing, whether by subdivision
or accretion, and all of a sudden a critical stage arises at which some of Figure 55
the units find themselves abruptly crowded inward, cut off completely
from direct contact with their former vital environment by an outer
layer of their fellows. The latter thereby acquire positions not only
geometrically intermediary, but functionally mediatory, between the
ambient medium and the now inner units. From then on, "inner" and
"outer" units are no longer alike. A monotonic group of equals has
become dichotomized into unequal sets. With the emergence of the Figure 56
 258 Within the Gates of Science and Beyond 259
One Plus One Does Not Equal Two

distinction between innerness and outerness, the 1 + 1 = 2 rule becomes An inorganic model of this process is, for instance, a sitting drop of
inapplicable.
mercury. Its convex, lens-shaped form results from equilibrium between
The train of events to follow such a "differentiation" of a radially opposing sets of forces-gravitation and adhesi~n, tending t~ spread the
symmetrical core-crust dichotomy is easy to envisage. Interactions mass, and cohesion and surface tension, tendmg to hold It together.
between the "outer" members and their newly established "inner" Disturb the equilibrium by cutting the liquid drop in two, and each half
neighbors would expose to another set of new conditions any fresh immediately restores its own equilibrium by assuming a convex
units arising subsequently in the intermediate zone between them, and lens-shape. But freeze the original lens-shaped drop solid before cutting,
hence call forth in them a third type of reaction. Moreover, polarized and then bisect, and each half will retain its former shape of half an
influences from the environment (e.g., gradients such as illustrated oblate; the dynamics that do the remolding in the liquid drop are still at
above for cell orientation) would impose an axiate pattern upon the work, but deprived of their free mobility, the elements can no longer
group. Thus would ensue a train of sequelae of ever-mounting,
yield.
self-ordering complexity. In all these steps, the fate of a given unit The example of twinning is just one illustration among m~ny for
would be determined by its response to the specific conditions the thesis that strict determinacy (or invariance) of a collective end
prevailing at the site in which it has come to lie, those conditions state is fully reconcilable with indeterminacy (or variance) in detail of
varying locally as functions of the total configuration of the system-its the component courses of events leading up to it-a thesis I ~ave tried
"field pattern," for short. This principle-long recognized empirically as to contrast with the basic reductionist doctrine that a determmate end
a basic criterion of systems but not always fully appreciated in its can only be reached as the blind outcome of a microprecisely
implications-is commonly referred to as "position effect." determined tandem chain of component microevents. This latter
The main point to bear in mind is that none of the component doctrine, "microscopic" and micromechanistic in the old sense, just is
members of the group, all erstwhile alike, can know their future courses not tenable in the light of facts unobscured by artificial blinders; and
and eventual fates in advance; can know whether they would become yet its popularity has grown steadily because of the indisputab~e proof
"inner" or "outer" or "intermediate." Nor does it matter for the that in the progress of science, as I said before, the artifact of
resulting pattern of the complex as a whole, as is best illustrated by the reductionist abstraction has had a most signal pragmatic merit. But the
process of twinning. By cutting in two the cluster of cells that time has come when we must check back with real nature to find what
constitutes an early embryo or an organ rudiment, one can obtain two we have missed by adopting the short-sighted view of close-range
fully formed embryos or two fully formed organs, the way the analysis as the sole legitimate approach to insight into nature. My early
sorcerer's apprentice, in trying to kill the water-carrying broom by introduction of the "field" concept into biology has aimed at no more
splitting it down the middle, got two busy whole brooms instead. What than at offering a semantic therapeutic against the spread of this
had been destined to form a single typical organism or organ has epidemic of myopia and constriction of the visual field, which leave~ so
yielded two instead, each half assuming the organization of a many burning problems in the life sciences unattended. The "fiel~" IS a
well-proportioned whole. In principle, we can now understand why: symbolic term for the unitary dynamics underlying order~d beha:lOr of
because bisection through the middle has resulted in "innermost" cells a collective, denoting properties lost in the process of Its phYSIcal or
coming to lie "outermost" again, whereupon the whole pattern of purely intellectual dismemberment. Being descriptive of a prop~rt! of
subsequent dynamic interactions has proceeded, reduced to half-scale, natural systems, it must not be perverted into a supernatural prmclple;
in harmonious proportions. (Of course, the individual parts can respond the study of those properties is, of course, an empirical task and not a
to their new local cues appropriately only if their original positions in
literary pastime.
the undivided framework have not already single-tracked them into If the young generation were only to realize the origin of the
courses unresponsive to the new demands.)
microdissectionists' claim for a monopoly of insight into nature, more
260
Within the Gates of Science and Beyond One Plus One Does Not Equal Two 261

of them mi~ht tur~ to problems now kept out of their purview. So let An experiment is motivated by our curiosity about the relationship
~e close with a brIef anamnesis of the prevailing conceptual deficiency between two phenomena of nature, A and B. We study them by
dls~ase. To me, .the crux see~s to lie not so much in a priori reasoning changing A from A' to A" and observe a correlated change of B' to B".
as m our practice of phrasmg experimental results in some sort of We then proceed to correlate the difference (A"-A') with the
shorthand language. I shall explain this on the schematic model of an difference (B"-B'). And basically this is all we can extract from the
experiment (Figure 57).
experiment. But this is not the point at which we commonly stop. We
usually go on to endow the differentials with an existence of their own,
dissected from the context from which they were abstracted in the first
place (bottom line in the diagram), and before we realize it, we have
personified them as "actors." Genes for the difference between a white
B' and a pink pea became simply genes for white and pink, respectively,
throwing the peaness into discard; the differences between integrated
brain functions before and after local lesions became transliterated to
domiciles for specialist subfunctionaries, as if the rest of the brain were
uninvolved; and so on.
In trying to restore the loss of information suffered by thus lifting
{!('
isolated fragments out of context, we have assigned the job of
B" reintegration to a corps of anthropomorphic gremlins. As a result, we
are now plagued-or blessed, depending on one's party view-with
countless demigods, like those in antiquity, doing the jobs we do not
understand: the organizers, operators, inductors, repressors, promoters,
regulators, etc.,-all prosthetic devices to make up for the amputations
which we have allowed to be perpetrated on the organic wholeness, or
to put it more innocuously, the "systems" character, of nature and of
our thinking about nature.
May my presentation have succeeded in documenting that party
lines drawn between emphasis either on the whole or on the parts are
based on the artifice of predilection, rather than on the realities of
nature. And may this realization find its way into human ecology,
particularly its political branch, to prove that society is not called upon
to choose between two extremes: either a license for anarchic random
excursions of its component individuals, or the enforced subordination
of individual members to a rigid group order dictated from above, but
that, as in all organic systems, order in the gross emerges, not only in
B"-B'
spite of, but as a result of, the interaction of free elements with
freedom in the small, restrained only by common purpose-or call it
program-and respect for nature, which after all, to speak in pre-
Galilean terms, abhors not only a vacuum, but disharmony.
Figure 57