Newton, Goethe and The Mathematical Style - VINE, Troy
Newton, Goethe and The Mathematical Style - VINE, Troy
Newton, Goethe and The Mathematical Style - VINE, Troy
the Mathematical
Style of Thinking
A Critique of Henri Bortoft’s
Taking Appearance Seriously
/ Troy Vine
I always carry Spinoza’s Ethics with me; he brought mathematics into ethics, as I did into
the science of colour [Farbenlehre]. That means there is nothing in the conclusion that is not
grounded in the premise.1
- Goethe
1. Introduction
J
ohann Wolfgang von Goethe’s approach to science is often characterized as holistic.
The popularity of this characterization is due, in no small part, to Henri Bortoft’s
influential interpretation based on the idea of two kinds of unity.2 Bortoft concludes
The Wholeness of Nature by contrasting what he refers to as Goethe’s “holistic” approach
to science with the “analytical” approach of experimental science (328–330). In his most
recent book, Taking Appearance Seriously, he contrasts the “concrete” nature of Goethean
science with the “abstract” nature of experimental science. Experimental science is abstract
because of the abstract nature of not only mathematics, but also of the mathematical style
of thinking that is reflected in a twofold experimental method.
Bortoft’s focus on the twofold method in experimental science, which proceeds from
experience to theory and from theory back to experience, brings out an important
aspect for understanding Goethe’s approach to science and its relation to experimental
science. I argue, however, that this shows not a difference between these two approaches,
I would like to thank Sonja Dorau, Philip Franses, Charles Gunn and Thomas Raysmith for
comments on drafts of this essay.
1. Goethe (2007, 621). My translation. Goethe made this comment in 1815 in conversation with the
German art historian Sulpiz Boisserée.
2. For an historical account of these two kinds of unity in Kant and their development in Goethe,
see Förster (2012).
[ 72 ]
but rather an important similarity which is often overlooked. The difference between
the two approaches cannot therefore be characterized by Bortoft’s distinction between
concrete and abstract, which is based on a distinction between the mathematical and the
hermeneutic styles of thinking. Using the example of colour, I show that this difference
is best captured by the distinction between relations that are necessary and those that are
contingent, which Bortoft calls “internal relations” and “external relations” respectively.
This twofold method is put forward in Aristotle’s Posterior Analytics, which was
rediscovered in the twelfth century and employed by scientists such as Robert Grosseteste
3. Bortoft also identifies a third cause of abstraction, namely mechanization. While I do not discuss
mechanization in this essay, it is central to understanding not only the difference between Descartes
and Newton, but also Goethe’s critique of Newton’s theory of colour.
[ 73 ]
and Roger Bacon in the thirteenth. They called the two stages “resolutio” and “compositio”,
which are Latin translations of the Greek “analysis” and “synthesis”. Crombie comes to the
conclusion that “the conception of the logical structure of experimental science held by
such prominent leaders as Galileo, Francis Bacon, Descartes, and Newton was precisely
that created in the thirteenth and fourteenth centuries” (3). As a result, “the history of the
theory of experimental science from Grosseteste to Newton is in fact a set of variations on
Aristotle’s theme” (318).
The general principles determined by analysis—the first part of the twofold method—are
not necessarily mathematical, and Francis Bacon gives “the most complete account of
the non-mathematical side of the theory of experimental science” (300). Bortoft, too,
includes Bacon among the scientists in whom “we find methodologically the same double
procedure that had been developed since Grosseteste” (31), and thus follows Crombie
in distinguishing between the application of the twofold method and the application of
mathematics itself.
Bortoft calls the twofold method the “mathematical style of thinking” because, “although
this double movement, from experience to theory and from theory to experience, is
formulated by Aristotle expressly for science”, it is “derived from the kind of reasoning
which he observed being practised by the mathematicians (30).4 The twofold method
has “the effect of shifting attention away from the phenomenon” (31), with the result that
“science becomes theory-centred instead of phenomenon-centred” (32). Moreover, “this is
particularly the case when mathematics begins to play a fundamental role in science” and
we “discover mathematical proportions and relationships in nature which lead us away
from the diversity of sensory appearances towards the discovery of a unity which is
more abstract” (32).
This idea of what Bortoft also calls an “abstract universal” led “to the remarkable idea that
there are universal laws of nature” (32). As we have seen, Bortoft distinguishes between
the application of the twofold method and of mathematics. Thus, while the abstract
universal is an expression of the mathematical style of thinking, it is not necessarily itself
mathematical. In a later passage on biology, for example, Bortoft contrasts the concrete
universal of Goethe’s archetypal plant (Urpflanze) with biologist Richard Owen’s “abstract
universal of a static generalisation” (83), which is a “minimal commonality from which all
the specialised organs required by actual living organisms have been excluded” (84).
The distinction between the abstract nature of both mathematics and the mathematical
style of thinking and the concrete nature of the sensory is a recurrent theme in Bortoft’s
book: “It is evident that, by its very nature, mathematics takes us away from the concrete
into abstraction. But this in itself does not necessarily undermine the value of the sensory”
(32). He also remarks that “although the mathematical style of thinking in physics leads us
away from the experience of the senses as such, there is no intrinsic reason why this should
make us think of the world as experienced through the senses as being inferior in any way
to the relationships in nature discovered by means of mathematics” (32). While in the first
passage Bortoft is contrasting mathematics itself with the sensory, in the second passage
he is contrasting the twofold method with the sensory. For, as we have seen, we do not
discover relationships in nature “by means of mathematics”, but by means of the twofold
method. This opposition between the sensory and both mathematics and the mathematical
style of thinking shows that Bortoft considers both to be abstract by nature, in contrast to
the concrete nature of the sensory.
Bortoft uses this distinction between the mathematical style of thinking and the sensory
4. I’m simplifying Bortoft’s account slightly because he does not seem to be aware that Aristotle
modelled what became the twofold method of resolutio and compositio on the twofold mathematical
method of analysis and synthesis.
[ 74 ]
to contrast Goethe’s approach to that of experimental science. He begins by stating
that Goethe “returned to the senses and put sensory experience first instead of the
mathematical” (53). And, in the context of plants, he remarks that:
The movement of thinking here is indeed very different from looking for
uniformities and commonalities in order to find a ‘general plan common to all
organs’, which is the approach so often wrongly attributed to Goethe. The dynamic
idea of the unity of nature that we find in Goethe is also very different from the kind
of unity we find in the universal laws of nature, which came from the mathematical
approach in science. (58)
Bortoft presents Goethe’s method as having two stages, which he calls “active seeing”
and “exact sensory imagination”.5 In the first stage, we put “attention into the sensory
experience itself, entering into the lived experience of sensory perception, so that rather
than just being ‘sensory’ in the empirical sense, it is better described as the ‘sensuous’
experience, or perception, of the phenomenon” (53). By becoming “aware of the sensuous
quality of each colour”, we transition from an empirical experience to sensuous experience
(54-5). Bortoft says of Goethe that by “redirecting attention into sensuous experience he
plunges into the sheer phenomenality of the phenomenon” (54).
[ 75 ]
blue, deeper blue, violet — but something more like ‘red-lightening—to—
orange—lightening—to—yellow’ as a dynamic whole, and similarly with
the darkening of blue to violet. There is a sense that the colours are different
dynamic conditions of ‘one’ colour. This dynamic quality gives us an intuition
of the wholeness of the colours at each boundary. This is not given directly
to sense perception, but appears when sensuous perception is sublimed into
intuition through the work of exact sensory imagination. In this way the
sensuous-intuitive mode of perception replaces the verbal-intellectual mode.
The colours are no longer thought of as being separate (verbal-intellectual)
but are experienced as belonging together (sensuous-intuitive). The way to
the wholeness of the phenomenon is through the doorway of the senses and
not the intellectual mind. We find there is the sense of a necessary connection
between the qualities of the colours at each boundary. It is not just accidental,
for example, that the order of the colours is red, orange, yellow — and not
red, yellow, orange — but is intrinsic to the colours themselves. (55–6)
Later in the book, Bortoft presents an example from mathematics to show “the difference
between the mathematical and the empirical” (158). His gives the description of the proof
that the sum of the interior angles of a triangle equals two right angles:
A proof which would be mathematically acceptable would be one that did not
involve measurement at all. It would be given entirely in terms of relationships
between the angles without any need to refer to the actual size of the angles in a
particular triangle. Consider any triangle ABC with angles a, b, and c (see Figure 1).
Extend the side BC [Bortoft means AB] into a straight line, and draw a line through
vertex A [Bortoft means C] parallel to this line (see Figure 2). Angle a equals angle
a' because they are alternate angles between parallel lines. Angle b equals angle b'
for the same reason. But angles a', b', and c must add up to 180° because they make
a straight line. Hence it follows that angles a, b, and c must add up to 180°. In such
a deductive proof we see that the angles of a triangle must add up to 180°. This is
entirely different from just saying that the angles of a triangle do in fact add up to
180°. It's not that they happen to do so — as if this were an empirical discovery —
but that they cannot not do so. (158)
a b
A B
Figure 1. Triangle ABC. (Bortoft 2012, 157)
[ 76 ]
C
a' c b'
a b
A B
Figure 2. Triangle ABC with side AB extended and parallel line passing through vertex C. (Bortoft 2012, 157)
Central to this description is the distinction between a contingent relation between the
angles, which we could determine empirically through measurement, and a necessary
connection: we see that the angles of a triangle must add up to 180°, not that they merely
happen to do so in this particular case. The nature of geometrical proof, then, is the seeing
of internal relations. To complete the proof, we need to add the stage that allows us to see
that alternate angles between parallel lines must be equal. To do this we can extend line AC
to see that angle a must be equal to angle a'' (see figure 3). Then, by rotating the line AC
about vertex C in our imagination, we can see that a'' must be equal to a' and thus that a
must be equal to a'. Similarly for b and b'.
C a''
a'
A B
Figure 3. Triangle ABC with side AC extended to show that alternate angles between parallel lines are equal.
In these two examples, there are no obvious criteria which allow us to apply the term
“abstract” to one and “concrete” to the other. The only criterion for applying these terms
would seem to be the movement from seeing external relations between empirical objects
[ 77 ]
to seeing internal relations between properties of objects (i.e. primary and secondary
qualities), which could be regarded as a process of abstraction. Bortoft remarks in a
footnote that:
Plato’s achievement was to show that what is truly mathematical does not depend on
working from sensory images of geometrical figures — for example, the discovery
that the sum of the interior angles of a triangle is equal to two right angles (180°)
does not depend on measuring the angles of drawn triangles, but follows directly
from the very idea of a triangle. (184)
The comparison above, however, suggests a different account. In the example of colour,
we work from a sensory image (we can consider exact sensory imagination to be sensory
because we “visualise what we have seen as exactly as we can”). As Bortoft characterizes
seeing internal relations not as an empirical experience but as a sensuous-intuitive
experience, Plato’s achievement could, therefore, be better described as showing a way of
working from sensory images that is not empirical.
Plato’s achievement, then, was to show that seeing external relations is distinct from
seeing internal relations, or, to put the same point differently, that empirical experience is
distinct from sensuous-intuitive experience. Even if it did “follow directly from the very
idea of a triangle” that the sum of the interior angles is equal to two right angles, the idea
itself consists of internal relations between geometrical elements (lines, points, etc.). The
empirical diagram on the page is not the concept of a triangle, but it does represent the
concept when we see the internal relations between the parts. The fact that Bortoft presents
a proof with diagrams suggests that, rather than being an unnecessary detour, working
from sensory images is necessary for grasping that very idea. The question, then, is not
whether we need to work with sensory images in geometrical proofs, but how we work
with them.
When we grasp the concrete unity of colours, “instead of abstracting unity from diversity,
we have the intuition that the diversity is within unity” (57). Triangles, however, are an
example of mathematics, and therefore, according to Bortoft, an abstract universal. Yet in
the geometrical proof above we do not abstract unity from diversity—this would only be
the case if we found empirically that the sum of the angles of a triangle are equal to two
right angles. Rather, as with the example of colour, we can grasp the triangle “dynamically
in its coming into being”. This can be made clearer if we move vertex C of triangle ABC
in our imagination: There is a sense that the triangles are different dynamic conditions
of ‘one’ triangle (substituting “triangle” for “colour” into the description of exact sensory
imagination above).
Let us consider the example of figure 4. How many triangles are there? If the concept
triangle is an abstract universal—i.e. formed by abstracting what is common to the five
figures—we must say that there are three triangles, because, as Bortoft notes, “all triangles
are three-sided” polygons (218) and only three of the five figures have this property in
common. However, if we grasp that the sum of the interior angles of a triangle is equal
to two right angles we can say that there are five triangles: we can see the straight line AB
as representing a triangle with one straight angle c and two zero angles a and b, as well as
the line AB with two lines extended at right angles from the ends as represent a triangle
with two right angles a and b and zero angle c. Moreover, we can see the five figures as five
representations of one triangle—i.e. the concrete universal—dynamically in its coming
into being.
[ 78 ]
C
C
C
A C B A B A B A B A B
Figure 4. Three triangles ABC, five triangles ABC, or one triangle ABC dynamically coming into being.
every possible triangle is subsumed in advance under the universal concept ‘triangle,
of which any triangle is therefore a particular instance. Everything is included in the
universal concept, so it is unthinkable that the universal itself could be enhanced by
any particular triangle. The movement is only from the universal to the particular
and never the other way round, so there simply cannot be any enhancement of the
universal by the individual case. (125)
According to Euclidian geometry, a finite line AB with two lines extended at right angles
from both its ends is not a triangle because parallel lines do not meet. But if we see
an internal relation between this figure and a triangle—e.g. by extending the vertex C
of the triangle in our imagination—, then we have an example of a particular triangle
that enhances the universal. Thus, not only is an “enhancement of the universal by the
individual case” conceivable, we have just done it!
The universal is enhanced, however, by abandoning the parallel postulate and thereby
opening up the possibility for entirely new kinds of geometry. If we abandon the parallel
postulate whilst retaining the possibility of measuring angles, there are two options. The
first option is to retain the idea that the sum of the interior angles is equal to two right
angles and extend Euclidean geometry to include ideal points. The second option is to allow
the sum to be less than two right angles, which gives hyperbolic geometry, or greater than
two right angles, which gives elliptic geometry (these two names are misleading because the
triangle remains a three-sided polygon with straight sides in non-Euclidian geometry).
This example shows that Bortoft’s static conception of mathematics is mistaken. Rather, as
Michael Beaney and Robert Clark have shown, the idea of seeing internal relations “sheds
light on the historical development of mathematical concepts” (2018, 133). This does not
mean, however, that we must give up the distinction between the abstract and concrete
universal; we just need to keep in mind that mathematical entities are not based on
abstracting commonalities from empirical objects, and therefore not abstract universals.
Bortoft’s examples of exact sensory imagination and geometrical proof show that there
is no essential difference between them. His distinction between the abstract nature of
mathematics and the concrete nature of exact sensory imagination is, then, a distinction
without a difference. We will now turn to his contrast between Goethe’s method and the
experimental method.
[ 79 ]
is created by Goethe’s view of the relation of his method to that of mathematics. In a
methodological reflection on his prism experiments, which we will consider in detail in
the next section, Goethe describes sensuous-intuitive experience as follows:
Bortoft’s contrast between Goethe’s approach to science and the mathematical style of
thinking is therefore at odds with Goethe’s own description of his method. In this section
we will assess Bortoft’s contrast by considering the methodology of Newton’s and Goethe’s
prism experiments.
The most detailed description Newton gives of the methodology of his prism experiments
is in an unpublished draft preface of the Opticks, which was first published in 1704:
As Mathematicians have two Methods of doing things which they call Composition
& Resolution & in all difficulties have recourse to their method of resolution before
they compound so in explaining the Phaenomena of nature the like methods are
to be used & he that expects success must resolve before he compounds. For the
explications of Phaenomena are Problems much harder than those in Mathematicks.
The method of Resolution consists in trying experiments & considering all the
Phaenomena of nature relating to the subject in hand & drawing conclusions from
them & examining the truth of those conclusions by new experiments & drawing
new conclusions if it may be from those experiments & so proceeding alternately
from experiments to conclusions & from conclusions to experiments untill you
come to the general properties of things [& by experiments & phaenomena have
established the truth of those properties]. Then assuming those properties as
Principles of Philosophy you may by them explain the causes of such Phaenomena
as follow from them: which is the method of Composition. (McGuire, 184–5)7
In this description, Newton is explicit about the relationship of his twofold method to the
mathematical method of analysis and synthesis: analysis determines general principles;
synthesis explains phenomena.
In the Opticks, Newton remarks that in the first book he “proceeded by this analysis to
discover and prove the original Differences of the Rays of Light in respect of Refrangibility,
Reflexibility, and Colour” (Newton, 405). He continues with the remarks that “these
Discoveries being proved, may be assumed in the Method of Composition for explaining
the Phaenomena arising from them” (Newton, 405). An example of composition is his
explanation of the rainbow. Newton’s prism experiments, then, are an example of the
mathematical style of thinking in experimental science and a good example of Bortoft’s
distinction between the application of the twofold method and the application of
mathematics. Mathematics can only be applied to colour once the general principle that
equates colour with refrangibility has been determined by analysis. The analysis itself,
6. I have modified Miller’s translation to render the German “Erfahrung” as “experience”, rather
than “(piece of) empirical evidence”.
7. The square brackets are Newton’s additions. For an overview of the history of these passages and
their relation to Newton’s methodology, see Shapiro (2004).
[ 80 ]
then, is not the application of mathematics to colour, but the application of the
mathematical method by analogy to determine general principles.
We will now turn to the method Goethe used in his prism experiments. A couple of
months after publishing the second part of his Contributions to Optics in 1792, Goethe
wrote a short methodological essay, quoted from above, which he later published with the
title “The Experiment As Mediator between Object and Subject”. In it, Goethe states that:
My intention is to collect all the empirical evidence in this area, do every experiment
myself, and develop the experiments in their most manifold variations so that they
become easy to reproduce and more accessible. I will then attempt to establish the
axioms in which the empirical evidence of a higher nature can be expressed, and see
if these can be subsumed under still higher principles. (1988, 17)
Here we can discern three stages: experimentation (variation of the experiments), seeing
internal relations (experience of a higher kind), and determining general principles. In a
short methodological essay called “Empirical Observation and Science”, written in 1798,
these three stages are summarized under the rubric of “empirical phenomenon”, “scientific
phenomenon” and “pure phenomenon” (1988, 25). These correspond to Bortoft’s three
kinds of experience that we saw in section 3, namely empirical, sensuous, and sensuous-
intuitive experience. However, while Bortoft describes sensuous experience as a plunge
“into the sheer phenomenality of the phenomenon” (54), Goethe’s describes scientific
phenomena as seeing relationships that are “fully perceptible” (1988, 14-5). Bortoft’s active
seeing and exact sensory imagination are the transitions from empirical experience via
sensual experience to sensuous-intuitive experience.
In general, events we become aware of through experience are simply those we can
categorize empirically after some observation. These empirical categories may be
further subsumed under scientific categories leading to even higher levels. In the
process we become familiar with certain requisite conditions for what is manifesting
itself. From this point everything gradually falls into place under higher principles
and laws revealed not to our reason through words and hypotheses, but to our
intuitive perception through phenomena. We call these phenomena archetypal
phenomena because nothing higher manifests itself in the world; such phenomena,
on the other hand, make it possible for us to descend, just as we ascended, by going
step by step from the archetypal phenomena to the most mundane occurrence in
our daily experience. (§175/1988, 194-5)
Goethe’s method proceeds from phenomena to general principles, which Goethe calls
“archetypal phenomena” (Urphänomene). Once these general principles have been
determined, they can be used to explain other phenomena. Goethe says that the principles
show themselves not to reason, but to intuitive perception. In other words, the principles
are seen, not merely thought. This corresponds to the idea of seeing connections and
Bortoft’s distinction between the verbal-intellectual and sensuous-intuitive mind.8
[ 81 ]
known phenomenon in a certain sequence so that we could determine the degree to which
all might be governed by a general principle” (1988, 48).
These passages show that Goethe is using the twofold method of analysis and synthesis
in his prism experiments. Thus, both Newton and Goethe are using the mathematical
method of analysis and synthesis by analogy. Goethe’s method, like Newton’s, is based on
the mathematical style of thinking. Bortoft’s contrast, then, between Goethe’s approach to
science and the mathematical style of thinking in experimental science is mistaken, and so
we need another distinction to capture this difference. In the next section, we will turn to
Bortoft’s distinction between external and internal relations by considering Newton’s and
Goethe’s prism experiments.
This kind of connection between the qualities of the colours is missing from the
Newtonian theory which asserts that light consists of colours which are separated
when it is passed through a prism. In this case there is no intrinsic necessity in
the order of the colours, but only an order that is imposed extrinsically by the
attribution of a wavelength to each colour. (56)
Bortoft’s contrast between Newton’s and Goethe’s account of colour is based on the
distinction between external and internal relations. We will develop this idea by
considering Newton’s and Goethe’s prism experiments.
In 1672, Newton published his “New Theory about Light and Colors” in the Philosophical
Transactions of the Royal Society.9 He begins by developing the following problem.
Having allowed a narrow beam of sunlight to enter his darkened room through a small
aperture in the window shutters and pass through a prism, he noticed that the image
that fell on the wall opposite was coloured rather than white and five times longer than
it was wide. He was able to calculate this difference to be much greater than Descartes’
theory of refraction could account for. Thus, he was able to show not only that there was
a hitherto undiscovered geometrical problem about light, but also that this problem was
bound up with the problem of colour. By combining the geometrical problem with the
chromatic problem in this way, a solution to the geometrical problem provides a solution
to the chromatic problem. Moreover, the solution to the chromatic problem is in terms of
geometry, rather than hypothetical corpuscles.
Newton’s solution is his experimentum crucis. Placing a board with a small aperture just
after the prism and a second about twelve feet away allowed him to select which part
of the coloured spectrum passed through the two apertures by rotating the prism. His
selection was refracted a second time by a prism placed behind the second aperture before
it fell onto the wall opposite (see figure 5). Newton found that light from the violet end
of the spectrum was refracted by a greater amount by the second prism than light from
the red end. As the path of the light remained the same (it passed through the same two
apertures), he came to his famous conclusion that “the true cause of the length of that
[original prismatic] Image was detected to be no other, than that Light consists of Rays
differently refrangible” (3079).
9. Cohen (1958, 47–59). The page number given for citations refers to the original publication of
Newton’s Letter.
[ 82 ]
Figure 5. Newton’s experimentum crucis. Light from the sun S passes through the aperture F in the window
shutters and the prism ABC to produce a spectrum on board DE. Rays from a region of the spectrum are
selected by aperture G in the board DE and aperture g in the second board de. The rays passing through
both apertures pass through the second prism abc and fall on the screen at point M. Different regions of the
spectrum are selected by rotating prism ABC (Newton, 47).
Newton uses the experimentum crucis to show that a geometrical property of a light ray—
the amount by which it is refracted—is related to the colour it produces on the screen: “As
the Rays of light differ in degree of Refrangibility, so they also differ in their disposition
to exhibit this or that particular colour”. Moreover, “to the same degrees of Refrangibility
ever belongs the same colour”, and “this Analogy ’twixt colours, and refrangibility, is very
precise and strict” (3081). Thus, Newton solved the geometrical problem by showing that
refrangibility is a property of light that is different for different kinds of rays, and solved
the chromatic problem by equating colour with refrangibility. Newton managed, then, to
give an account of colour in terms of refraction, and an account of refraction in terms of
geometrical rays.
Newton’s account of colour, then, leaves a gap in our understanding; for while we can
see that refrangibility is the immediate (proximal) cause of colour, we can still ask for a
further (remote) cause to explain why a given degree of refraction causes the particular
colour that it does.10 As a result, despite being a geometrical explanation rather than a
mechanical explanation—in Newton's terminology a theory rather than an hypothesis—, it
nevertheless opens the door to mechanical explanation. It is thus not surprising that beside
the geometrical account of light and colour in the Opticks we find a mechanical account,
although the two kinds of explanation are kept distinct.
We will now turn to Goethe’s prismatic experiments. In 1791, Goethe published the
first part of the Contributions to Optics. Despite being his first publication on colour, it
contains his most perspicuous presentation of prism experiments. The presentation is
based on observing black and white patterns through a prism.11 It begins by showing that
a homogenous white or black card seen through a prism remains unchanged (§41) and
that a boundary between light and dark is necessary for prismatic colours to appear (§42).
This is followed by a number of complex forms presented that produce different colours,
including black with white bands that produce Newton’s spectrum (§44).
In order to “analyze these wonderful appearances”, Goethe decomposes Newton’s
10. For an account of proximal and remote causes in Newton’s method see Ducheyne (2012, 18–47).
11. The first part of Contributions to Optics is in Goethe (1951, 6–37). My translations.
[ 83 ]
spectrum, which is produced by two boundaries between light and dark, into two spectra
each of which is produced by a single boundary (§45). A horizontal boundary of black
above white produces a distinct band of red above a distinct band of yellow, whereas a
horizontal boundary of white above black produces a distinct band of blue above a distinct
band of violet (§47-8). Goethe then provides a card with both situations next to each other
so that the two edge spectra can be compared: blue appears opposite red and violet appears
opposite yellow (see figure 6). This “shows that the colours do not follow one another,
but “oppose one another” as “two opposing poles” (§§55, 72). Goethe has shown that the
prismatic colours are governed by a principle of polarity.
Figure 6. Goethe’s illustration of the two edge spectra produced by looking at a boundary between black and
white through a prism. Contribution to Optics, Card 14 (Goethe 1951, Plate VIII).
To produce Newton’s spectrum, the two boundaries are brought together by viewing the
black card with a horizontal white band first close up and then moving it further away. The
two edge spectra separated by the white boundary come together, and green appears where
the blue and yellow bands overlap (see figure 7). The next card is white with a horizontal
black band, and the situation is reversed. This time magenta, or what Goethe calls “peach
blossom”, appears where the violet and red bands overlap (see figure 8). By moving the
cards still further away, “the mixtures peach blossom and green […] totally extinguish the
colours of which they are composed” (§59).
Figure 7. Goethe’s illustration of the spectra produced by viewing a black card with a horizontal white band
through a prism. Green appears where yellow and blue overlap. Contributions to Optics, Card 9
(Goethe 1951, Plate V).
[ 84 ]
Figure 8. Goethe’s illustration of the spectra produced by viewing a white card with a horizontal black band
through a prism. Goethe has not reproduced the magenta band that appears where violet and red overlap.
Contribution to Optics, Card 10 (Goethe 1951, Plate V).
Goethe has demonstrated, then, that the Newtonian spectrum can be produced by
combining two edge spectra. In addition to producing the familiar Newtonian spectrum,
Goethe uses the principle of polarity to produce a complementary composite spectrum in
which all the colours of the familiar Newtonian spectrum are replaced with their opposite,
or complementary colour. Moreover, the inner two bands of the edge spectra, which
overlap to produce green or magenta, disappear completely, leaving just three coloured
bands: red, green and violet for the Newtonian spectrum; blue, magenta and yellow for the
inverted spectrum. It thus appears as if the two inner bands of colour mixed to produce a
new colour.
Bortoft suggests, as we saw in section 3, that the order of the prismatic colours “is intrinsic
to the colours themselves” (56). We can apply this idea of internal relations between
colours to understand Goethe’s account of Newton’s spectrum. As we have seen, Goethe’s
account of the two composite spectra is based on the idea of colour mixing. Therefore,
there are two parts to an account based on an internal relation between colours. Firstly, we
must show that there is an internal relation between the colours of the edge spectra, and,
secondly, that there is an internal relation between the two colours that overlap and the
colour they produce. As Bortoft’s account presented in section 3 only addresses the first
part, we have not yet seen that the colour produced when the two interior colours of the
edge spectra overlap is internally related to the two interior colours.
Goethe’s account is based on the polarity of light and dark. This refers to the polarity not
only of white and black, but also of light colours and dark colours. We saw, in figure 6,
not only that white is opposite black, but also that a light colour is opposite a dark colour.
As the relation of light and dark is an internal relation, we can use it to see that Goethe’s
account is based on an internal relation between colours. We also saw that the edge
spectra appear at a boundary between light and dark. They consist of a light colour (yellow
or blue) next to white and a dark colour (red or violet) next to black. Thus, both edge
spectra have the structure: white, light colour, dark colour, black. The one edge spectrum
is therefore the opposite of the other in terms of the relation of light and dark. Thus, the
internal relation that hold between white and black also holds between the two colours
of the edge spectra: Yellow is lighter than red, blue lighter than violet; conversely, red is
darker than yellow, violet darker than blue. Figure 6 shows that there is an internal relation
between the colours of the edge spectra.
We have also seen that when two light colours overlap in the Newtonian spectrum, a dark
colour is produced. Conversely, when two dark colours overlap in the complementary
spectrum, a light colour is produced. Thus, two kinds of mixing occur, which are polar
with respect to light and dark: a mixing that lightens two dark colours when they overlap,
[ 85 ]
and a mixing that darkens two light colours when they overlap. Figures 7 and 8 show that
there is an internal relation between the two colours that overlap and the colour they produce.
These two kinds of mixing are usually referred to as additive and subtractive mixing.
However, this conception is misleading if one thinks of it as the addition or subtraction of
coloured lights. An account in terms of rays would be a causal explanation, and therefore
an empirical account based on external relations. Rather, when referring to a mixing of two
colours, Goethe has in mind a mixing of the property of colour that is analogous to adding
and subtracting the property of number. The statements “1 + 1 = 2” and “2 – 1 = 1” are not
empirical statements about objects, but logical statements about numbers. Similarly, the
statements about the mixing of colours by overlapping are not empirical statements about
coloured lights, but logical statements about colours.12
We are now in a position to derive the form of the Newtonian spectrum from the polarity of
light and dark. Because it is composed of two edge spectra whose light colours combine to
produce a dark colour, it must consist of five coloured bands, starting with a dark colour and
alternating between a light colour and a dark colour. We have thus given an account of the
form of the Newtonian spectrum in terms of the internal relation of light and dark.
A comparison between Newton’s and Goethe’s prism experiments in terms of external and
internal relations brings out an important difference: in Newton’s approach, we are able to see
a relation between an angle of refraction and a particular colour, but it is only in Goethe’s that
we are able to see an internal relation between the colours themselves. For Newton, the order
of the colours appears contingent, but Goethe shows that it is necessary. Put another way, in
Newton’s theory it is conceivable that the order of the prismatic colours could be different, in
Goethe’s it is not. Thus, despite Newton’s approach being geometrical, this comparison shows
that Goethe’s approach is closer to mathematics, because it is based on internal relations.
In the last section we saw that both Newton and Goethe use the twofold method to determine
general principles, and then use these principles to explain other phenomena. In this section
we have seen that Newton uses analysis to determine an external relation between the
refrangibility of a light ray and the colour it causes, and then uses synthesis to explain why
sunlight creates a coloured image when passed through a prism. Goethe, on the other hand,
uses an analysis of prism experiments to determine internal relations between prismatic
colours, which are expressed in the principle of polarity. He then uses synthesis to explain the
composition of Newton’s spectrum. As in the last two sections, Bortoft’s distinction between
concrete and abstract seems inapplicable here. Bortoft’s distinction between external and
internal relations, on the other hand, brings out an essential difference between Newton’s and
Goethe’s account of colour.
6. Conclusion
We began by considering Bortoft’s account of the history of experimental science and the
supposed abstraction due both to the application of mathematics and to the mathematical
style of thinking reflected in the twofold method. Juxtaposing Bortoft’s example of exact
sensory imagination with his example of geometrical proof showed that the difference
between them cannot be captured by the contrast of concrete and abstract, because
both examples are based on seeing internal relations between properties. Similarly, our
juxtaposition of Newton’s and Goethe’s methodological descriptions showed that both
Newton and Goethe used analysis to determine general principles and synthesis to explain
phenomena. Goethe’s approach cannot, therefore, be considered concrete in comparison to a
supposedly abstract mathematical style of thinking.
12. For an account of additive and subtractive mixing in relation to Goethe’s prism experiments, see
Wilson (2018).
[ 86 ]
Our comparison of Newton’s and Goethe’s prism experiments, on the other hand,
demonstrated that while Newton shows that there is an external relation between prismatic
colours and refrangibility, Goethe shows that there is an internal relation between the
prismatic colours themselves. Goethe demonstrated, then, that the order of the prismatic
colour is not contingent, but necessary. Our investigation shows that Boftoft’s distinction
between internal and external relations is able to capture the difference between Newton’s
and Goethe’s approach to colour. In addition, it shows that rather than representing a
break with experimental science, Goethe’s approach to science expresses the diversity
within the unity of the mathematical style of thinking.
References
Beaney, M. A. & R. Clark (2018) “Seeing-as and Mathematical Creativity”. In B.
Harrington, D. Shaw & M. Beaney, eds., Aspect Perception After Wittgenstein. Seeing-As
and Novelty, Routledge.
Bortoft, H. (1996) The Wholeness of Nature. Goethe’s Way of Science, Floris Books.
Bortoft, H. (2012) Taking Appearance Seriously. The Dynamic Way of Seeing in Goethe and
European Thought, Floris Books.
Cohen, I.B. (1958) Isaac Newton’s Papers & Letters on Natural Philosophy, Harvard
University Press.
Crombie, A.C. (1953) Robert Grosseteste and the Origins of Experimental Science. 1100–
1700, Oxford University Press.
Ducheyne, S. (2012) The Main Business of Natural Philosophy. Isaac Newton’s Natural-
Philosophical Methodology, Springer.
Goethe, J.W. von (1951) Beiträge zur Optik und Anfänge der Farbenlehre. 1790–1808, ed. R.
Matthaei, Böhlaus Nachfolger.
Goethe, J.W. von (1988) Scientific Studies, ed. and trans. D.E. Miller, Suhrkamp.
Goethe, J.W. von (2007) Zur Farbenlehre und Optik nach 1810 und zur Tonlehre, ed. T.
Nickol, Böhlaus Nachfolger.
McGuire, J.E. (1970) “Newton’s ‘Principles of Philosophy’: An Intended Preface for the
1704 Opticks and a Related Draft Document”, British Journal for the History of Science 5,
178–86.
[ 87 ]