Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116

Contents lists available at ScienceDirect

Studies in History and Philosophy of Biological and

Biomedical Sciences
journal homepage: www.elsevier.com/locate/shpsc

Justifying molecular images in cell biology textbooks: From

constructions to primary data
Norberto Serpente
Department of Science and Technology Studies, University College London, Gower Street, London, WC1E 6BT, UK

a r t i c l e i n f o a b s t r a c t

Article history: For scientic claims to be reliable and productive they have to be justied. However, on the one hand
Available online 16 September 2015 little is known on what justication precisely means to scientists, and on the other the position held by
philosophers of science on what it entails is rather limited; for justications customarily refer to the
Keywords: written form (textual expressions) of scientic claims, leaving aside images, which, as many cases from
Cell studies the history of science show are relevant to this process.
Textbooks
The fact that images can visually express scientic claims independently from text, plus their vast
Molecular images
variety and origins, requires an assessment of the way they are currently justied and in turn used as
Justication
Pedagogy
sources to justify scientic claims in the case of particular scientic elds. Similarly, in view of the
Image-based archives different nature of images, analysis is required to determine on what side of the philosophical distinction
between data and phenomena these different kinds of images fall.
This paper historicizes and documents a particular aspect of contemporary life sciences research: the
use of the molecular image as vehicle of knowledge production in cell studies, a eld that has undergone
a signicant shift in visual expressions from the early 1980s onwards. Focussing on textbooks as sources
that have been overlooked in the historiography of contemporary biomedicine, the aim is to explore (1)
whether the shift of cell studies, entailing a superseding of the optical image traditionally conceptualised
as primary data, by the molecular image, corresponds with a shift of justicatory practices, and (2) to
assess the role of the molecular image as primary data. This paper also explores the dual role of images as
teaching resources and as resources for the construction of knowledge in cell studies especially in its
relation to discovery and justication. Finally, this paper seeks to stimulate reection on what kind of
archival resources could benet the work of present and future epistemic historians in particular those
interested on the role of images as sources of training and knowledge production in scientic disciplines.
2015 Elsevier Ltd. All rights reserved.

When citing this paper, please use the full journal title Studies in History and Philosophy of Biological and Biomedical Sciences

1. Introduction However, what justication means for scientists and how pre-
cisely they justify the images they use as sources to produce new
For scientic claims, such as hypotheses and theories, to attain knowledge in different scientic elds and disciplines are topics
the status of scientic knowledge they have to be justied.1 that remain under-examined by philosophical and historical
E-mail address: n.serpente@ucl.ac.uk.
studies of science.
1
In epistemology, a long dated assumption is that knowledge is justied true In analysing cases of scientic justication philosophical studies
belief. This tripartite assumption, which was rst credited to Plato (427-347 BCE). Gail have customarily focused on the written form (textual expressions)
(2003), Bonjour & Sosa (2003), pp. 1, and found a modern expression, in the work of of scientic claims, a situation that has resulted in a neglect of the
the philosopher Alfred Jules Ayer (1910e1989, 1956), pp. 31e35, despite genuine
role of images in this process. In view of the increased recognition of
critics on its insufciency Gettier (1963), pp. 53e54, is considered by and large, by
epistemologists as roughly correct, Neta & Pritchard (2009), pp. 5, in that it remains the capabilities of images for promoting particular epistemologies,
an important bedrock assumption in epistemology, philosophy of science and science for acting as text-independent tools for visual thinking, and for
itself. Justication, because of its multiple meanings and associations with, for
instance, validation, verication, truth, demonstration, reliability, explanation, etc.,
has remained a highly debatable issue in epistemological discussions Alston (1989),
Neta & Pritchard (2009), Moser (1996), Landesman (2002), Musgrave (1993).

http://dx.doi.org/10.1016/j.shpsc.2015.08.007
1369-8486/ 2015 Elsevier Ltd. All rights reserved.
106 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
embodying scientic claims,2 this lack of critical examination of how
scientic images are justied has become untenable. This is espe-
cially so when considering the relentless proliferation of their
diverse forms alongside novel pedagogical trends in scientic dis-
ciplines in general and bio-disciplines in particular throughout the
late twentieth century and up to the present. Moreover, due to this
proliferation of visual expressions it is unclear how scientists epis-
temologically differentiate them in particular concerning the
important distinction identied by Bogen and Woodward (1988)
between data and phenomena (see also Woodward, 1989).
These authors proposed a distinction between data that is, all those
observable instrument-derivable inscriptions (Latour & Woolgar,
1986) obtained in experimental instances and non-explained by
the theory under test, and phenomena, a feature of the world,
usually unobservable, which is created from those instrument
derivable observations and is used as explananda for the theory
under test. The question is then in which side of this key philo-
sophical distinction scientists locate the different kinds of images
they work with.
Besides these two important questions, there is that of the
relationship between pedagogy, discovery and justication, and to
what extent all these epistemic activities could be of any value for
the constitution of image-based archives. Images play a key role in
training future generations of scientists and, in this regard, the
textbook emerges as a particularly unexplored source in the his-
toriography of science. Textbooks, and the role of their illustrations
in justifying scientic discoveries and transmitting them to
younger scientists, deserve examination, particularly for those
scientic elds that are highly dependent on images.
One such scientic eld is cell studies,3 for it is a eld that has
undergone a major quantitative and qualitative visual trans-
formation between the 1950s and the 2000s. In effect, during that
period cell studies have undergone a representational shift entail-
ing a superseding of images obtained with microscopesdimages
traditionally conceptualised as extension of naked eye vision and as
primary data; by molecular images, that is pictorial representations
constructed from the outputs (primary data) of one or more set of
instruments other than microscopes aimed at describing the mo-
lecular interactions assumed to be at the basis of anatomical and
functional changes in cells (Serpente, 2011a, 2011b). This shift to-
wards the prevalence of constructed images invites reection in
several fronts. First, the way these two kinds of images are justied;
second, how scientists categorise them, for instance, whether
constructed images, such as molecular images are regarded as data
or phenomena; third, how the justication process works on the
making of textbooks; fourth, and last what role if any pedagogy
plays for justication.
2
The examples here are numerous, but the following would clearly highlight the
point: a) the case of Alfred Wegener drawings on the movement of continents to
visually express the theory of continental drift. Robin (1992), pp. 210. b) The series
of drawings used to represent the phenomenon of cell division of pre-reproductive
cells (meiosis) to visually express the Mendelian theory of inheritance. For more
examples on the inference of theories from images and the construction of phe-
nomena out of data see the cases discussed by Brown. Brown (1996), pp. 250e268.
In Baigrie (1996). A typical example of images as independent of textual expres-
sions are the drawings of Leonardo Da Vinci. For cases of images promoting
particular epistemologies see Zampieri, Zanatta, and Bonati (2012), pp. 121e144. In
Fangerau, Chhem, Muller, & Wang (2012).
3
The term cell studies attempts to embrace the history of studies of cells from
the formulation of the cell theory in the 1840s to the present regardless of the
changes in name that the discipline underwent, rst in the 1960s from cytology to
cell biology, and second in the 1980s from cell biology to molecular cell biology.
Cell studies also embrace disciplines involving the study of cells interacting such as
plant and animal development and immunology.
The aim of this paper is to study the process of justication
behind the molecular image in cell studies and with it, to create
ground for reection on the importance of images as archival
sources for further epistemological enquiry. The pictorial work
of two textbooks that have proved foundational for cell studies
is analysed here: The Machinery of Life (TMoL) by David Goodsell
(rst published in 1983) and Molecular Biology of the Cell
(MBoC), by Bruce Alberts, Dennis Bray, Julian Lewis, Martin Raff,
Keith Roberts and James D. Watson (rst published in 1983).
The objective of this analysis is to explore both the question of
what counts as data and phenomena, and the position that
conceptualises textbooks, in contrast to scientic papers, as
sources for more established and hence more reliable forms of
knowledge.
All in all this paper will argue for a complex picture in the
production of microscopic and molecular imagery in cell studies, a
picture where apart from the conventional set of criteria estab-
lished by philosophy of science for the acceptance of scientic
claims, pedagogy and other factors play a crucial role. By assessing
images as primary sources of knowledge, this study highlights the
importance for archivists and historians of cataloguing images,
including the hidden questions and constrains that stand behind
their design. As such, and in line with other contributions to this
special issue, it addresses the difculties in interlinking archival
images and text and advocates for a productive crosstalk between
archivists and historians on how to create resources from where
future generations of academics could rescue the visual craft-like
aspects of scientic practices that are more difcult to reconstruct
from written sources.4
2. Discovery/justication and the visual nature of scientic
claims
Of particular relevance for the history of justication in sci-
ence was its demarcation from discovery, a demarcation credited
to Hans Reichenbach (1891e1953), a key member of the logical
empiricist school of philosophy.5 Reichenbachs idea was to
separate the rational steps based on formal logic and inductive
and deductive arguments (the relevant) from the psychological,
sociological and historical aspects (the anecdotic) when granting
new proposed hypotheses or theories the status of knowledge. In
consonance with the critical view of Kuhn and others on this
strict demarcation, this study assumes that, rather than being
4
On images and text as sources for archives see de Chadarevian, The future
historian: Reections on the archives of contemporary life sciences, 2016. On the
process of establishing and cataloguing an archive see Shaw, Documenting geno-
mics: Applying archival theory to preserving the records of the Human Genome
Project 2016. On the necessity of interactions between historians and archivists see
Garcia-Sancho, The proactive historian: Methodological opportunities presented
by the new archives in modern genomics, 2016. On the hidden and informal ac-
tivities that doing science entails and the difculties in documenting them see
Aicardi, Francis Crick, cross-worlds inuencer: Reections on the archives of
contemporary life sciences, 2016.
5
Reichenbach (1938), pp. 5e8 and 381e384. As remarked by Hoyningen-Huene
(1987), pp. 502e503, implicit and/or related versions of the distinction can be
traced back to Whewells The Philosophy of Inductive Sciences, which was published
in 1847, and also to other members of the logical empiricist school during the late
1920s, as well as to Karl Popper in 1935 (translated in English in 1959). The view of
Whewells philosophical work as a forerunner and himself as an advocate of the
distinction between both contexts has been contested by Jutta Schickore. Schickore
& Steinle (2006). The discovery/justication relationship is a theme that has kept
attracting intellectual interest throughout the years. See, for instance: Hanson
(1960), Kordig (1978), Nickles (1980). Hoyningen-Huene (1987 and 2006),
Schickore & Steinle (2006). A typical argument of the relevance of justication
over discovery found in science is that if a key idea/theory X has not been proposed
by scientist A in year 1 it would have been discovered by scientist B or C in year 2.
 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
sequential with discovery preceding justication (as granted by
logical empiricism), they instead overlap and coproduce each
other.6
At present, the position held by the vast majority of scientists
on justication, is that for scientic claims to be justied and
thus acceptable as genuine knowledge claims by other scientists,
1) they have to derive from experimental data, 2) they should be
able to be put forward as premises for new experiments, 3) they
should account for the new observable data to emerge from
those experiments, and 4) they can predict new phenomena from
those data.7 Of importance is the fact that, although stripped
from its originally mandatory logical arguments, justication has
survived as a key normative ideal in science, serving to allegedly
avoid the unwanted entry of subjective bias in research. More-
over, even though at present logical empiricism has, with the
advent of the historical turn in philosophy of science, lost its
appeal,8 its claims on the prioritisation of justication over dis-
covery have endured not only in philosophy of science, but also
in science.
A salient consequence of the historical prevalence of justica-
tion over discovery in philosophical and scientic outlooks is that of
the neglect of images as sources of knowledge independent of text.
This consequence has been grasped by Richer who, echoing Gieres
original suspicion,9 argued that this neglect of pictorial means in
philosophical studies is part of the legacy left by the logical em-
piricists, who [ . ] abstracted from pictorial means and set their
focus on language (Richter, 2014, pp. 58e60). This neglect of im-
ages is not exclusive to philosophy of science. Concerning history of
science, Daston has recently remarked: [ . ] how blind historians
of science once were to anything but words: scientic texts were
6
Kuhn implicitly recognised the distinction in the introduction of The Structure of
Scientic Revolutions Kuhn (1970), and endorsed the overlap between them in The
Essential Tension. Kuhn (1977), pp. 326e329. For further analysis see Hoyningen-
Huene (1993), pp. 245e252. For an empirical case study of their overlapping
taken from the history of molecular biology see Schaffner (1974).
7
This outlook derives from both informal discussions and interviews the author
has with cell biologists on how they justify ndings in their day to day practice.
Although in philosophy of science this outlook is more nuanced and is always under
renewed scrutiny, it is one on which most philosophers of science would agree
upon. This scientic outlook could be associated with reliabilism, that is, knowledge
that depends on a reliable process or method; in the case of science, the scientic
method. The scientic take on justication also relates to what is known in epis-
temology as the externalist theory of knowledge Lehrer (2000), pp. 17, in which the
connection of a scientic claim (belief) with reality (what is going on the experi-
mental system) is tested by experimentation and the observed consequences from
these experiments can be then compared with other evidence provided by other
observations and further experimental results. In essence this is the so called
hypothetic-deductive model, which in a nutshell consists of a cycle of action
comprised of the following steps: a) gather data, b) propose a claim accounting for
those data, c) deduce new observational predictions from the claims, d) set
experiment to test the claims. Adapted from Godfrey-Smith (2003), pp. 236. As
Kuhn has remarked, to these justication criteria can certainly be added those of
consistency, accuracy, scope, simplicity and fruitfulness Kuhn (1977), pp. 321e322).
8
The received view is that logical positivism was a dominant outlook in phi-
losophy of science throughout the 1940s and the 1950s until the early, mid 1970s,
when the potential contained in the work of Thomas Kuhn (1970) began to be
realised, with this being a watershed event for the further loss of appeal of logical
positivism and the concomitant opening of the historical turn in philosophy of
science. This outlook on the decline of logical positivism at the hands of Kuhn has
been contested from many angles. Firstly, it has been argued that logical empiricism
has never enjoyed such a period of intellectual supremacy, for their key ideas have
never been fully accepted by other philosophers. Richardson (2007). Secondly, its
demise has been attributed to other factors than the decline of logic and the rise of
an historical outlook in philosophy of science, such as the anti-Communist climate
that reigned in the US during the 1940s and 1950s, which was very hostile to the
key members of the logical empiricist movement. Reisch (2005).
9
He afrmed that In the framework of logical empiricism [ . ] there can be no
fundamental role in science for non-linguistic entities like pictures or diagrams
Giere (1996), pp. 270.
107
purely textual; when we came to an image (a drawing a graph, a
table, a diagram, a photograph, it was all one), we just ipped the
page (Daston, 2014, pp. 319). Moreover, she nds the use of ex-
pressions such as reading images and visual literacy as clear ex-
amples of this text-centered approach in image analysis (Daston,
2014).
I would like to argue that what has aggravated this neglect is
that, rstly, in most occasions images have been conceptualised
as pedagogical illustrations-as subsidiary learning tools sup-
porting text content- and secondly, to the fact that pedagogy has
traditionally been considered as irrelevant to epistemology.10 All
in all then it should not surprise us that under this regnant
text-centered and pedagogy devaluated scenario, the way
scientic images are justied and how they relate to data and
evidence, have not received the attention it deserves.
An alternative perspective however, one giving a far more
preponderant role to images in the production of knowledge,
began to emerge in the last 20 years by works exploring the
commonalities between art and science.11 This body of scholar-
ship argues for a subtle and complementary relationship be-
tween text and image, with one becoming preponderant over the
other depending on the context (Manghani, Piper, & Simon, 2006,
pp. 169). More importantly, what these studies also suggest is
that there are similar circumstances in science to those
commonly assumed in art, in which images become more ef-
cient than words to perform knowledge, in many cases attaining
an almost complete independence from them. Converging with
this view, several historical studies have emerged, highlighting
the multifaceted role of scientic images for the production of
knowledge.12
The role of images for knowledge production is enhanced by
their differential role as thinking tools, situation that is epitomised
in Martin Kemps assertion that for the artist as for the scientist
every act of looking has the potential to become an act of analysis
(Kemp, 2000, pp. 3). Many examples from well-dened historical
periods also reinforce his claim.13
A particular case of images combining the role of thinking
tools and embodiment of theory, which because of their nature
(cells and molecules) take us closer to the focus of this paper are
those created in the 19th century by immunologist Paul Ehrlich
(1854e1915) on antibodies. Art historian James Elkins has argued
that Ehrlichs pictorial depictions of antibodies at a time when
their structure was not known and hence their existence doubted
were essential for the development, in the short term, of his
immunological theory (side chain theory) and, in the long term,
10
This position also relates to the residual situation imposed by the former
philosophical analysis of science which by focussing exclusively on theory change
has abstracted knowledge production from its practices.
11
The list is not exhaustive but the following works are representatives: Elkins
(1995, 1999), Galison & Jones (1998), Gooding (2004), Kemp (2000, 2006). The
work of Kepes (1956) was a sort of anticipation.
12
For general works see the many essays in Baigrie (1996) and in Pauwels (2006).
As an example of images embodying theories James Brown has suggested, using
cases from mathematics and thought experiments from physics, how both theo-
rems and theories can be inferred and their demonstrations followed from images
Brown (1996), pp. 250e268. Concerning the role of images in supporting episte-
mology, it has recently been revealed how pivotal was the iconography of wax
models of smallpox as used by Luigi Sacco, an Italian physician and surgeon from
the early nineteenth century, to support his medical knowledge, which was largely
based on empirical observation, experimentation, and therapeutic intervention
rather than on theory Zampieri et al. (2012), pp. 121e144.
13
Among them for instance is that of Robert OHara concluding that for an
adequate establishment of order for the existing diversity of living forms during
the 19th century many concepts, were best expressed diagrammatically than in
writing OHara (1996), pp. 164.
108 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
of the core of what is currently accepted as the norm in immu-
nology.14 While the material nature of antibodies was uncertain
Ehrlichs imagery was justied not only because it materialised
and rendered visible these hypothetical structures, but also
because they acted as pedagogical tools to be used as heuristic
devices for subsequent experimentations.15 Ehrlichs image-
based epistemology/pedagogy, and in particular, the use of a
justication process based on rendering visible hypothetical
structures, is one that has proved extremely productive in the life
sciences and as such endured the passing of time. This image-
based epistemology/pedagogy was profusely used during the
twentieth century by Linus Pauling (1901e1994) and others, and
manifested in the pictorial work displayed in cell studies
textbooks.
Some questions arise here: if images are able to embody sci-
entic claims and scientic claims to be valid need to be justied,
how are images justied? How does a cell biologist go about
determining whether a molecular image displayed in a textbook or
a scientic article is a reliable source of knowledge when planning
new experiments? And in view of the distinction between data and
phenomena: on which side of the divide cell biologists locate the
different types of images they work with? Given that the survival of
an image as a knowledge source is largely shaped by scientists
accepting it and conferring it a sound epistemological status, the
answer to these questions is highly relevant for historians. To
address them this study will focus on two types of images: the
microscopic and the molecular, images which have been at the
epicentre of the shift that the discipline of cell studies underwent in
the early 1980s.16
3. Microscopic images in cell studies and their justication
Microscopic images are the product of magnifying devices such
as optical and electron microscopes.17 The building of knowledge
on cells from the times of cell theory in the 1840s to the present
has hinged on a process of justication of images obtained with
these two types of devices. Two aspects are worth noticing when
considering this process of justication. Firstly, even though
anatomical and/or functional claims about cells are based on the
14
Elkins (1999), pp. 38e39. Ehrlichs theory of embodied images was not exempt
of controversy. As amply documented by Cambrosio, Jacobi, & Keating (1993),
Ehrlichs images were highly criticized by the French immunologist Jules Burdet,
who labelled them as illusory explanations. It is not an exaggeration to say that
this dispute between Ehrlich and Bordet constituted a dening moment for the role
of images in biology, for it unfolded at a time when the ontological status of an-
tibodies was purely hypothetical and technologies to determine it were non-
existent (absence of instrument based data).
15
Cambrosio et al. (1993), pp. 682e694. This is an idea originally proposed by
Gooding for the work of Michael Faraday. See: Gooding (1990).
16
Scientic images are of a varied nature and they of course include more than
microscopic and molecular variants such as diagrams, graphs, drawings, photo-
graphs, microphotographs, models, etc. The focus in this study on the microscopic
and the molecular image is because they have been at the centre of the visual shift
cell biology gone through in the early 1980s. Serpente (2011a, 2011b), and because
as we will see in what follows, different conventions and justicatory practices are
at play for the transformation of the instrument output into valuable (meaningful)
visual data in both of them.
17
The images that these two instruments produce, although from a different
nature, share some common features. They derive from the condensation (focus) in
a visual plane of the scattered particles of energy (light or electrons respectively)
that have passed initially through a chemically treated and stained sample, then
subsequently pass through a magnifying lens and are registered by a camera as a
photograph (after reection of the electrons in a uorescent screen for the case of
the electron microscope). For a brief, but solid account on key aspects of the history
of the optical microscope see Hughes (1959), pp. 1e28. For the case of the electron
microscope see Rasmussen (1997).
interpretation of what microscopic images show (the image itself
is conceptualised as primary data), the claims are usually
expressed in a textual form with images downgraded to the role of
illustration of the point being made in the text. Secondly, the
justication of microscopic images has been closely linked to is-
sues of reliability and standardisation of both sample preparation
techniques and technical improvements of instrumentation.18 In
other words the tuning of appropriate conditions for the xation
and staining of cells,19 alongside the nding of solutions to the
many technical problems that microscopes presented, such as for
example, in the case of the optical microscope, the existence of
chromatic aberrations, was key for justifying the use of the rst
microscopic images of cells to build working hypotheses of their
structure and functions20.
This close association between justication of factual claims
and the justication of methodological standards could be better
appreciated from a survey of the production of images in electron
microscopy during the 1940s and 1950s.21 Currently taken for
granted, the justication of the rst electronic images of the
substructure of cells entailed many interpretative dilemmas. As
the controversy that took place over the internal structure of the
mitochondria between Fritiof Sjostrand (1912e2011) and George
Palade (1912e2008) in the 1950s suggests, the process of learning
to see and a quick consensus on what was to be seen was crucial
for cytologists involved (Bechtel, 2006, pp. 192e209; Rasmussen,
1997, pp. 124e152). In other words, the history of microscopy is
also a history of how the senses have been schooled and
extended (Daston & Lunbeck, 2011, pp. 2). In fact, it is not an
exaggeration to afrm that achieving consensus on the determi-
nation of what counted as a genuine and reliable image and what
was a ctitious one, a process in which pedagogy played a key
role, was one of the main concerns on research laboratories in
cytology and virology between the 1940s and 1950s.22 Two
pedagogical strategies were pivotal in facilitating both the justi-
cation of images delivered by the electron microscope and in
schooling the senses in order to make the images meaningful.
18
The justication of factual claims goes hand in hand with the justication of
methodological standards, a key component to render them trustworthy for stu-
dents and researchers. For a brief, but solid account of the history of the optical
microscope images see Hughes (1959), pp. 1e28.
19
Fixation of biological samples such as cells and/or tissue samples involves the
use of chemicals such as aldehydes and alcohols in order to prevent the denaturing
of its internal proteins and thus preserve their integrity for microscopic observa-
tion. Staining is used to increase the contrast among the different cell structures
from the samples. The problem of contrast for the case of the optical instrument
was solved by the development of a new generation of dyes and for the electronic
instrument by using (spray on samples) of heavy metals salts, such as Osmium.
20
As for example the case of chromatic aberrations in optical microscopy which
were solved by the development by Ernst Abbe (1840e1905) in 1886 of apochro-
matic lenses. Chromatic aberration was one of the main suspects for the production
of artifactual images in optical microscopy. For the case of electron microscopic
images the main ones were, on the one hand, the incapacity of the beamed elec-
trons to penetrate the thick specimens in use by cytologists at the time, and on the
other, the instruments incapacity to produce an acceptable level of contrast.
Rasmussen (1997), pp. 27.
21
The images created by the electron microscope far from being straightforward
were highly controversial and criticised on many fronts. See for instance the
Hacking/van Fraassen controversy in: Hacking (1985), (it includes van Fraassen
answer to Hacking). See also the many technical and conceptual issues raised in
Hillman & Sartori (1980). See also Rasmussen (1997).
22
As Palade vividly put it when asked about the early days of electron micro-
scopy: it was like opening a huge [ . ] uncharted province, of a huge equivalent of
a forbidden city, or forbidden continent. Cited in Rasmussen (1997), pp. 232. My
argument is based on the type of papers published at the time and on the dis-
cussions that followed every time a new sub-cellular structure was proposed to
exist. For an extensive discussion on these issues see Rasmussen (1993, 1995, 1997).
 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
Firstly, the selection of cell images that presented visual elements
close to those in images obtained through optical microscopy, a
process that I dened elsewhere as the transfer of iconicity
(Serpente, 2011a) and secondly, the use of paired representations,
a practice that emerged in the early 1950s- and is still amply used,
consisting in pairing an electromicrograph with a drawing or di-
agram in order to facilitate its interpretation (Lynch, 1990, pp.
157e168). Finally, a key event for the justication of the electronic
images, an event that is hailed as a watershed for the history of
cell studies, was their use to identify the subcellular structures
present in the different fractions obtained from the ultracentri-
fugation of cell extracts involved in specic biochemical functions,
such as the case of mitochondria and the Krebs cycle (Bechtel,
2006).
4. Molecular images in cell studies
The microscopic image is not the only kind of image that has
contributed to produce the present knowledge on cells. The
molecular image has also played a part in this process, mainly as
the consequence of the relentless molecularisation that cell
studies and other disciplines such as biochemistry have under-
gone from the 1930s onwards (De Chadarevian & Kamminga,
1998; Kay, 1993). Molecular images are of various types. The
particular kind of molecular image that this study focuses on
corresponds to models presenting arrays of different molecules
(proteins and in many cases nucleic acids) interacting, aiming to
explain changes on cell states and diverse cell functions such as
movement, cell differentiation, secretion, and tissue and organ
development. (Serpente, 2011a). An important category of mo-
lecular images of this nature, because of its extensive use in
scientic articles and textbooks, is that portraying signal trans-
duction processes (Heldin & Purton, 1996; Gomperts, Kramer, &
Tatham, 2002). A key point to recall here is that such molecular
images describing the multiple interactions of proteins involved
in the transmission of a signal from a cell surface receptor to the
cell nucleus, derive from the output of instruments, which do not
create images of cells directly like a microscope. Their production
hinges complex processes of translations of data from non-
optical instruments alone or in combination, into constructed
pictorial representations.23 Many aspects of the process of con-
struction of molecular images have been exposed (Knorr-Cetna &
Amann, 1990), yet the way they are justied and their ambivalent
positioning with regard to the data/phenomena divide, are two
issues that have so far remained unexamined. This is what the
next section of the paper will now focus on.
5. Justifying molecular images in textbooks
Key to assessing the process of justication of molecular
images in cell studies is to consider one of the main media
through which they are communicated-namely textbooks. This
will allow us to evaluate what justication means to scientists,
how they justify the molecular images they produce, where they
23
Input from other techniques is often used. Although impossible to give a full
description of all the techniques used a list to help the reader if in need of detail is
given in what follows: Genetic studies, antibody mediated Immunoprecipitation/
Western blotting (IP/WB), yeast two hybrid system, optical (immunouorescence
proteineprotein co-localisation studies), electronic microscopy and X-rays (studies
of atomic density) are all involved in the production of molecular imagery. Signal
transduction studies for instance are by and large based on the interpretation of
traces (black dots) left by radioactively labelled macromolecular complexes
detected in autoradiography lms, assumed to be the result of proteins interacting
in cells.
109
position them inside the data/phenomena distinction and to
appreciate the intimate relationship between justication, dis-
covery and pedagogy at play during the production of images.
Given that textbooks are mainly read by younger generations of
student scientists, the justication of their images is a
crucial duty to be undertaken by their authors. In this regard,
textbooks and the particular communication process that is
established around their images become a crucial and rather
unexplored source to study the history of contemporary life
sciences.
I will rst focus on The Machinery of Life (TMoL), a book pro-
fusely illustrated with hand drawings and computer generated
images of molecules (Fig. 1), which was rst published in 1983
with its most recent edition appearing in 2009 (Goodsell, 1983/
2009). Its author is David S Goodsell, currently an Associate
Professor of molecular biology at the Scripps Research Institute
(Florida, USA). Soon after earning his PhD on DNA structure from
the University of California at Los Angeles in 1987, and motivated
by his robust interest in art Goodsell became a pioneer in the use
of computer graphics in biomolecular structural research, with
his work featuring in the covers of several magazines, scientic
journals and in the column Molecule of the Month from the
Protein Data Bank website.24 Goodsell started his career as a
molecular designer by producing a series of small articles
(Goodsell, 1991, 1992; Olson & Goodsell, 1992), in which he dis-
played the molecular design style that he would later expand in
TMoL. Goodsells work was motivated by the lm Fantastic
Voyage25 and builds from the tenets of haptic visuality, that is a
tactile way of seeing based on the viewers bodily every-day
experience rather than what is seen with the eyes. Much in
line with this, Goodsells molecular images are an open invitation
to viewers to experience by themselves the internal molecular
composition of cells by assuming different multi-focus positions
in an act where discovery and learning become almost
indistinguishable.26
In the preface of TMoL, he wrote:
Imagine that we had some way to look directly at the mole-
cules in our bodies, perhaps with an x-ray microscope or an
Asimov style shrinking and enlarging machine (unfortunately,
neither is currently feasible). Think of the wonders we could
witness rst hand: antibodies attacking a virus, electrical
impulses shooting down nerve bers, proteins building new
strands of DNA. Many of the questions puzzling the current
cadre of biochemists would be answered at a glance. But the
microscopic world of molecules is separated from our
everyday world by an insurmountable, million-fold difference
in size.27
24
This is an archive containing the 3D-D structure of proteins. See, 2014 See:
http://www.rcsb.org/pdb/101/motm_archive.do (Accessed last 13th March 2014).
25
He made this explicit in Goodsell (1992), pp. 457. The screenplay of Fantastic
Voyage was written by Harry Kleiner and David Duncan and it was released as a
lm in 1966. Pfeiffer (2014). Briey, it describe the tribulations of a group of sci-
entists when they are placed inside a submarine, which is reduced to the size of
bacteria, and injected into the bloodstream of a colleague who held critical and
secret information, but that has suffered a stroke produced by a blood vessel cloth
in the brain. Equipped with laser guns their aim is to x the brain clot in a race
against time before their return to the original size.
26
The art historian Alois Riegl (1858e1905) argued that this kind of images for
simultaneously seeing and touching were part of Egyptian art. Iversen (1993).
Concerning bodily experience, electron microscopy captured the public imagination
in 1960s in the US as a practice that allowed viewing the minute by transporting the
viewers body into it. Rasmussen (1997), pp. 222e256.
27
Goodsell (1983). From the preface (no pages number).
110 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
Fig. 1. Goodsells molecular images: A) Global view of an entire Mycoplasma mycoides cell. B) An internal view of a portion of the bacteria Escherichia Coli showing: (from down-right
corner towards the up-left corner), the nucleoid region containing DNA (yellow) and the enzyme DNA polymerase replicating it (orange), the cytoplasmic area with ribosomes
(purple), small strands of messenger RNA (white) and transcriptional enzymes (blue), and the cell membrane studded with transmembrane proteins and a agellar motor crossing
the membrane (green).
Source: http://mgl.scripps.edu/people/goodsell/illustration/public
Goodsell is keen to portray his work on molecular images as
both realistic, and naturalistic.28 In his view his depictions are
realistic, in that they are mirror images of the world, accurately
representing how crowded by molecules cells are (Goodsell, 1991,
pp. 205). Besides, he claims that these images provide the viewer/
witness with a clear picture of the interior of a living cell that
shows the average distribution of molecules at the proper scale
(Goodsell, 1991, pp. 203). Similarly, he sees his work as naturalistic,
in that it attempts to match the observers direct experience of
what is being represented,29 by showing what a viewer might see if
the subcellular structure were magnied to a scale large enough to
make individual macromolecules visible (Goodsell & Johnson,
2007, pp. 2761). He also justies his work as both necessary and
unique because in absence of an adequate technology it lls an
otherwise invisible visual void that exists between the too coarse
view left by the images obtained with the electron microscope and
the too ne one obtained with X-ray crystallography (X-RC) and
nuclear magnetic resonance (NMR). (Goodsell, 1991, pp. 203).
To defend the realism and naturalism of his images, Goodsell
constantly stresses the use of experimental scientic results in his
work. In effect, he uses a combination of different sources of
instrument-derived data, all entailing their own justication pro-
cesses. The creation of images, apart from some drawing conven-
tions, follows three successive steps (Goodsell & Johnson, 2007, pp.
2761). He rst uses electron micrographs as sources of visual
empirical data to contextualise the cellular localisation of the mole-
cules or molecular complex chosen to be depicted. Goodsell trusts
these micrographs as reliable data to dene subcellular structure and
thus takes their justication for granted. In the second step he looks
28
The position assumed by Goodsell is not without controversy for it is highly
debatable that these categories are attainable. Also because it assumes on the one
hand that the referent is rendered as such by technical mediation and on the other
that the referent even if it is invisible (molecules) could be compared with its
representation. On this front, as Pauwels points out, justication problems arise
when the referent is postulated and the evidence for its existence is reached
through an indirect process. Pauwels (2006), pp. 9.
29
Kemp (2014), pp. 344. These images because of their condition of eyewitness
pictures arguably gave viewers a deep sense of trust in them (Kemp, 2014), pp.
345).
for images related to the 3-D structure of the individual molecules
that have been selected, typically supplied by data from NMR and X-
RC. Goodsells third and last step is that of synthesis, in that he locates
all the molecular structures collected in step two in the place pro-
vided by the rst contextual image, yet magnifying them.
Goodsell also justies the scientic worth of his images by
selectively using a series of what he calls metaphors, or concepts
(Goodsell, 2005). Goodsells metaphors either reveal the internal
structure of macromolecules or depict not just the intracellular
environment where those macromolecules are located, but also the
place where the imaginary observer sees it. Among the former are
bond diagrams, which better represent the covalent structure of a
molecule; space lling representations, which better portray the
contact distance between atoms (the one he more frequently uses);
and ribbon diagrams, representing more accurately the folding
protein chains (Goodsell, 2005, Fig.1, pp. 348). Two key metaphors
of the second type are those of immersive approach and cross
section, both referring to where inside the cell he wants to locate
the viewers gaze. Whilst for the rst the camera is placed within
the scene so that the viewer is surrounded by molecules, for the
second the camera remains outside it, so that the iconicity of
electron micrographs allows the display of large eld of molecules
and entire molecular processes.30
From our previous discussion it emerges that the justication of
molecular images in TMoL and Goodsells work in general depends
on a highly constructive process were scientic data and artistic
leeway (artistic license in his own denition), delicately mesh
alongside pedagogy. By artistic license Goodsell and Johnson refer
to the right measure of artistic input required to transform scientic
data in images for further scientic thinking, decisions that, in their
view have to resist editorial aims to produce ashy illustrations with
only hypothetical elements (Goodsell & Johnson, 2007). This peda-
gogic component is in fact central to sustain the tight connection
between data and artistic leeway (Goodsell & Johnson, 2007, pp.
2759), and key to the choice of placing the viewer/learner within the
eld of subcellular structures and molecules, a move that not only
enhances comprehensibility but also empowers observers.
30
Goodsell (2005), pp. 352e353.
 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
Fig. 2. Roberts molecular images: A) The so-called walking protein, showing its different conformational shapes designed to express the idea of an energy induced molecular
movement.
Source: Figure 3e63, MBoC, 1989. Courtesy of Professor Keith Roberts. B) Cascade of proteineprotein interactions based on energetically induced molecular shape changes designed
to express the cellular phenomenon of cell survival. Source: Figure 15e60, MBoC, 2002. Courtesy of Professor Keith Roberts.
Some scientists think that Goodsells molecular images can help
elicit further experimentation (another key tenet of justication).
David Rudner, a microbiologist from Harvard Medical School,
thinks that by being incredibly instructive in both mnemonic and
predictive terms these images are able to stimulate both general
discussion and foster hypotheses.31 Goodsells own view is that
scientic illustrators [ . ] not only depict current scientic
knowledge, but also contribute to hypotheses as they form
(Goodsell & Johnson, 2007). This capacity of Goodsells illustrations
highlights the tight relationship between discovery, justication
and learning, and the key role of images and textbooks to histori-
cally capture the knowledge production and justication process in
cell studies.
The other book of relevance to the arguments advanced in this
paper is Molecular Biology of the Cell (MBoC), rst published in 1983,
and currently in its sixth edition (2014). The textbook was the result
of an initiative led by James Watson who by the mid-1970s
perceived that cell studies were too anatomical and too factual
and that these features, in view of the state of molecular studies of
cells at the time, called for a change (Serpente, 2013). MBoCs rst
edition thus constituted a synthesising effort to encapsulate the
latest experimental results on eukaryotic cells, such as the mech-
anisms of their genetic regulation and intracellular signalling, re-
sults that were relentlessly being produced in cell biology labs
worldwide but that were not covered by any existing textbook. This
feature, alongside the books abundant and imaginative use of
molecular images and many other characteristics of its production,
made it both an embodiment and a hallmark of the molecularisa-
tion of cell studies that began in the late 1970s (Serpente, 2011a).
31
http://www.asbmb.org/asbmbtoday/asbmbtoday_article_print.aspx?id13702
(Accessed 20 May 2014).
111
The person who created the imagery of MBoC is Keith Roberts, a
former Professor at the John Innes Centre, an independent research
centre of reference in plant biology at Norwich, UK. Before being
recruited by James Watson as a co-author for MBoC, Roberts
completed a PhD in biochemistry in 1970 at the University of
Cambridge, while retaining from his secondary school years an
active interest in art.32 Concerning his pictorial work, before pro-
ducing the imagery for MBoC, he created the images for Watsons
Molecular Biology of the Gene (1965) and many other textbooks over
the years, among them John Kendrews The Thread of Life (1966), as
well as the renowned Albert Lehningers A Short Course in
Biochemistry (1973). He also produced many cartoons presenting
molecular interactions for the scientic journal Trends in
Biochemistry (TIBS).
MBoC displays different kinds of molecular images, but two
stand out as the textbooks trademark ones (Fig. 2). Firstly, visual
metaphors such as the walking tooth aimed at conveying the idea
of molecular movement by changes in protein conformation
induced in turn by the changes in the molecule energy states,33 and
the safe cracker, aimed at expressing the view that proteins can
work in a coordinated and mechanical manner. Secondly, those
pictorial representations aimed at conveying the different cyclical
states of molecular complexes involved in the accomplishment of
cellular phenomena by using amorphous bodies of different size,
32
The book: The new landscape in art and science (1956) by Gyorgy Kepes, a
Hungarian painter, art theorist and professor at the College of visual arts at Harvard,
exerted a deep inuence on Roberts work on molecular images. Serpente (2011a),
pp. 135e136.
33
Alberts et al. (1983), pp. 137e138. The walking tooth and the safe cracker are
the names the authors used among themselves to refer to these single protein and
protein complexes respectively.
112 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
which change shape as a consequence of their interaction.34
Together, the two kinds of visuals stand as the best examples of
the molecular conceptualism (the capacity for images to embody
conceptual ideas on cell functioning independently of text) that
Roberts wanted to achieve for his images. Molecular conceptu-
alism, one of the many novelties used in the making of the rst
edition of MBoC, which was maintained in its successive editions,
proved essential for its pedagogical reach. In effect MBoCs image-
based molecular conceptualism not only served to straightfor-
wardly transmit to students themes that as the years passed by kept
growing in complexity, but also to rmly establish the nowadays
taken for granted view of molecular explanation at the base of all
cellular phenomena (Serpente, 2011a, 2011b, 2013).
All the molecular images displayed in MBoC are justied by an
extensive reliance on experimental data. For this, Roberts uses
images from electron microscopes as well as many other non-visual
techniques such as X-RC, NMR and other immune-biochemical
methods. As sources for these data he used the original experi-
mental output provided by scientists, from unpublished results to
published scientic papers. Roberts constructed his own molecular
images by visually contextualising those scientic data within a
exible artistic input, and with the aid of what he termed prag-
matic rules of depiction.35
The making of MBoC and of its images in particular was very
closely linked to the development of ongoing experimentation in
labs worldwide. This dynamic process worked as follows. When the
MBoC authors evaluated and discussed the unpublished experi-
mental data they received from colleague scientists, they produced
hypotheses of putative molecular interactions. They would then ask
the data providers to perform further experiments to test these
hypotheses (Serpente, 2011a, pp. 124). The close dynamics between
scientic claim making, ongoing lab experimentation and students
feedback, not only made of MBoC an original textbook, but one that
challenges the traditional view of pedagogy, as peripheral to the
work done in laboratories, (Gaster, 1990, pp. 431), one that also
shows the weakness of the discovery/justication distinction.
Moreover this dynamics also challenges the traditional distinction
between textbook science, one which contains more stable and
consensually accepted work and frontier science, that of reports
and scientic articles where only a more argumentative, specula-
tive and less proven work features.36
In many ways Roberts process of creation of molecular images
for cellular events resembles that of Goodsells, in that both aim to
give priority to a productive mix of primary data, and artistic
creativity, alongside pedagogy for knowledge claims. It is worth
bearing in mind that their pictorial approaches, despite their reli-
ance on new technologies, are far from novel. They clearly belong to
a tradition of arguing and witnessing with images initiated by Paul
Ehrlich and continued by Linus Pauling (1901e1994), a tradition
informed by structural and physical chemistry, and framed by
pedagogical strategies in which hypothetical mechanisms are
materialised (Cambrosio, Jacobi, & Keating, 2006) and thus trans-
formed in epistemic things, that is objects that will generate
34
Alberts et al. (1983), pp. 747. These images belong by and large to the eld of
signal transduction and are not exclusive to Roberts.
35
Which are the following: I) Illustrations need to be unied and consistent
throughout the book. II) A gure should tell only one story, as economically as
possible. III) Scale has to be consistent throughout all the illustrations. IV) Use
colour only to contribute to the overall feel and coherence of a text. V) Create
gures in ways that they can tell more than one story. Roberts (2003), pp. 439e440.
36
This view holds that research published in scientic papers does not qualify as
scientic knowledge, it is instead information that has been widely available. Only
undergraduate textbooks qualify as scientic knowledge that has wider currency,
Bauer (1986), (1992, pp. 47).
questions (Rheinberger, 1997). However, one important difference
between Goodsells and Roberts work is that whilst Goodsells aims
at representing cells as realistically and naturalistically as possible
(as crowded molecular environment), the focus in Roberts images
remains explanatory and conceptual. In other words, by pointing
to the idea of visually summarising molecular mechanisms, Rob-
erts molecular images stress the idea that interactions matter more
than the precise form and size of molecules. That said, in the 2002
fourth edition of MBoC an interesting pictorial synthesis occurred.
Rather than using amorphous globular forms to represent protein
movement such as transmembrane ion channels, Roberts began to
use instead images of molecules derived from NMR and X-RC data
in his depictions, thus fusing his molecular conceptualism with
Goodsells molecular realism.37 This is a move that Goodsell might
well have welcomed for, in an interview, with clear reference to the
imagery contained in the rst editions of MBoC, he declared: When
other people have attempted to draw cells at the level I am doing,
they simplify things by using circles and triangles to point areas of
uncertainty. Ive resisted that because it breaks the illusion that this
is a photograph or a direct representation (Miller, 2002, pp. 1).
Some thoughts concerning the positioning of microscopic and
molecular images on the data/phenomena distinction emerge by
reecting on the production of molecular imagery in textbooks. To
begin with, it transpires that microscopic images qualify as data
because, rstly, they are taken as the primary sources from which in
part molecular images are constructed, secondly they are consid-
ered as the original sources against which to evaluate the hypo-
thetical claims expressed either textually or by the molecular
image, and thirdly they are taken as sources which are arguably not
theory laden and thus have an immediate factual validity. Of
course the status of microscopic images as data and hence as
warrantors of epistemic reliability is one that takes for granted the
many instances from the historical process of image production in
optical and electronic imagery in which their role as sources of
direct evidence for justifying hypothesis, that is their epistemic
reliability, has been questioned.38
A rst glance at molecular images suggests that, because of their
nature as theory laden pictorial constructions, they qualify as
phenomena. Nevertheless, reecting on the way they work in
textbooks, it results that they can also act as data. In effect, mo-
lecular images, due to the epistemic claims that they embody, can
be used as premises for further hypotheses and experiments. That
said, molecular images do not possess the immediate factual val-
idity that characterises microscopic images. Their theoretical val-
idity is constantly deferred because their conrmation depends on
ongoing experimental processes. At this point, it is worth noting
that this categorisation of microscopic images as data and of mo-
lecular images as both phenomena and data is not exclusive to the
authors of textbooks; it also corresponds with the way bio-
scientists doing cell studies conceptualise them. A clear manifes-
tation of this is how they use these images in their day-to-day
practice: paper sketches of molecular images are pinned above
every bio scientist bench and desk as altogether evidence of pre-
vious ndings, acting as route planners to put forward new hy-
potheses on cell function, and resources for planning new
experiments to test these new hypotheses.
37
A clear example is that of the sarcoplasmic reticulum calcium pump. Alberts
et al. (2002), pp. 626. Compare it with the cytosolic calcium pump. Alberts et al.
(1994), pp. 743.
38
Think of all the verication processes at play in optical and microscopic imagery
to produce reliable, non-artifactual images and all the controversial events that
populate the history of traditional microscopically based cell biology on micro-
scopes aberrations and use of staining technique with dyes.
 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
The other conclusion that emerges from our previous analysis
concerns the criteria used for the validation of microscopic and
molecular images in textbooks. It seems that the classical scientic
justication process is not strictly at play for the production of
molecular imagery as it is the case for microscopic imagery. Mo-
lecular images do not have, as microscopic images do, a set of strict
criteria for their empirical validation such as, for instance, specif-
ically designed controls to determine their veracity. A key factor for
the validation of molecular images, apart from their correspon-
dence with data, is their pedagogical worth. I would like to argue
that in attaining this pedagogical worth the validation process at
work for molecular imagery is closer to that applied to classic
artistic visual expressions. This, in my view, is manifest in the
overriding concern for attaining naturalism away from explana-
tion in Goodsells case or in the making of a conceptual visual story
away from realism in Roberts case.
To sum up, it seems that molecular imagery follows a justica-
tion process that is more relaxed and mixed than the one classically
used for written expressions of scientic claims. Molecular images
are justied, by a) microscopic images, b) the internal consistency
with the epistemology from which they originate, c) their corre-
spondence with the accepted translational process from the tech-
nologies that produce them, d) their capacity to elicit new
experiments in other research groups and nally, e) the decision of
the thought collective (Fleck, 1979, pp. 38e51), that accept them as
pedagogical sources.39 No doubt then that molecular imagery plays
a central role in the production and reproduction of knowledge,
because of its pedagogical worth and their capacity to both
constitute phenomena and give rise to new experiments.
6. Images and pedagogy, a subtle link
Despite their diversity and their multiple origins, scientic im-
ages in cell studies share a common feature, their pedagogical
worth. In this respect, what matters most is their efciency and
insightfulness for transmitting knowledge. I would like to argue
that through their use they also have the capacity to elicit a tacit
replication of the process of development of scientic claims,
including their justication. This capacity of images to tacitly
replicate the process of development of scientic claims especially
in classrooms situations is one of the aspects that Lynch refers to
when he argues that visualisation is far from limited to practices
that allow something to be seen, observed, or depicted (Lynch,
2014, pp. 324). This last aspect, to which we will pay attention in
what follows, is one that although arguably recognised has received
scant attention, chiey because as discussed before, images in the
bio sciences have been strictly perceived as auxiliaries to text.
Conceptualising pedagogy exclusively as a key to the spread of
knowledge after it has been produced and validated, rather than
placing it within the process of knowledge production (discovery), is
a misleading assumption inherited from logical empiricism and one,
which most current bio-scientists still adhere to. This assumption
however is not exclusive to logical empiricism. Its survival has been
helped by the ambivalent position that one of its main challengers,
Thomas Kuhn (1922e1996), adopted on pedagogy and knowledge
production. When answering his critics about his views on the
impossibility to set apart discovery from justication Kuhn created a
new context, that of pedagogy, which in his view differs almost as
39
Also, as I previously discussed, Serpente (2011b), molecular images also became
legitimate tools for knowledge production and knowledge claims from their ca-
pacity as symbols to contain iconic and indexical forms as well as from their em-
bodiments of values from the culture from which they emerge. See also Perini
(2010).
113
much from the context of justication as it does from that of dis-
covery (Kuhn, 1977, pp. 327). Kuhns ambivalence derives from the
fact that despite explicitly locating pedagogy well apart from both
contexts, he simultaneously acknowledged the relevance of the
relationship between scientic training (teaching) and scientic
practice (knowledge production). As Kaiser points out, the notion
that solving-problem activities somehow replicate the original work
done by scientists when advancing claims is implicit in Kuhns work.
(Kaiser, 2005, pp. 394e395).
As I have shown elsewhere (Serpente, 2011a, 2013), epistemic
developments in cell biology concomitant to changes in the pro-
duction of textbooks during the early 1980s (Gaster, 1990), de-
velopments from which MBoC is constitutive, show how close
pedagogical practices are to the justication of new knowledge. Of
course, pedagogical methods are not exclusive to textbooks. The
use of images in textbooks and in classroom teaching practices in
biosciences are about taking students into simultaneously exer-
cising the gaining of knowledge about a subject and applying
rational judgement to evaluate the content and validity of such
knowledge. As Schefer has noticed, learning to understand a
subject is learning to know that subject (Schefer, 1965, pp. 17)
which is another way of saying that any pedagogical process entails
meeting the requirements of justication. In other words, the ex-
amination by students of images with an embodied epistemic and
pedagogical content entails exercising the tenets of justication.
This is manifest in current educational practices in science, which
consist in engaging students in activities resembling both experi-
mental and justication practices. As shown by Smith and
Blankinship (2000), students in classrooms learn by arguing and
debating over differences on image data rather than just absorbing
uncritically the content of images (Smith & Blankinship, 2000, pp.
764). What is more, image thinking in textbooks and pedagogical
settings runs much in line with the work done in the lab where the
examination and re-examination of images after the proposal of
hypotheses result in a cyclical activity where imagery acts simul-
taneously as method for problem solving and as evidence.
These less solemn and more routine activities in which re-
searchers informally debate and agree on paths to follow have been
traditionally excluded from the idea of doing science (Aicardi, 2016).
This has resulted in the neglect of historically interesting lines of
work and the exclusion of potentially insightful scientic sources.
Pedagogy and imagery are clear examples of this, especially given
that images, as historical materials, require a different treatment
from text and are routinely discardeddor disconnecteddfrom ar-
chives (de Chadarevian, 2016; Shaw, 2016). By looking at the role of
images within a broader conception of doing sciencedone that does
not exclude trainingdthese valuable sources and practices may be
preserved for future historians.
Of course, the intertwinement between teaching and knowl-
edge production is not a new phenomenon. There is evidence of its
emergence in 19th century German universities, where teaching in
the elds of embryology and medicine was simultaneously con-
ducted in lecture and demonstration rooms known as spectato-
riums (Anderson & Dietrich, 2012). It can be argued that the
practices deployed in spectatoriums were one of the rst cases
where the traditional division between pedagogy and experimental
science was dissolved.40 As a concluding point, it is clear that
40
This complex intertwinement between experimentation, knowledge produc-
tion and pedagogy is not exclusive to the biomedical sciences, for, as suggested by
the work of Andrew Warwick, it has also been at play in the mid nineteenth century
development of mathematical education in Cambridge. Warwick (2003), and dur-
ing the process of theory change in twentieth century chemistry. Soon Park (2005),
pp. 287e322.
114 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
pedagogy and experimental science remain intertwined more than
a hundred years afterwards, even if scientists and scholars, because
of their endorsement of the discovery/justication dichotomy
continue to separate them articially in their writings.
7. Concluding remarks
This paper has explored the relationship between imagery
production and justication in the case of cell studies, a eld that in
the last fty years has undergone a signicant change in the
quantity and more centrally in the quality of its imagery content,
with molecular images prevailing over microscopic ones. I have
suggested that images of a molecular nature have been exempted
from questions concerning epistemic justication both because
they have been assumed to be the direct consequence of theory
(where theories are justied by written statements) and because
they have been conceptualised as pedagogical devices and peda-
gogy has not been traditionally considered as having relevance for
epistemic justication.
As a result of this, the sources that historians traditionally
consult to reconstruct the development of cell studiesdand more
generally contemporary biosciencedare dominated by written
documents and a narrow conception of what doing science entails.
Exploring the role of images within a broader denition of science-
in-the-making (Aicardi, 2016) may thus provide a more inclusive
framework to evaluate both activities and sources in historical
analysis. A conception of doing science that does not articially
separate pedagogy and experimentation may lead to a consider-
ation of teaching materials as valuable entry points into the
workings of 20th century life sciences. I have shown this by ana-
lysing the complexity of the production and justication of images
in two cell studies textbooks.
The pedagogical straightforwardness of MBoC, a key textbook
for the molecularisation of cell studies from the early 1980s on-
wards, has hinged upon a well-organised laboratory-based web of
interactions that Roberts and the other authors were able to
establish with colleagues worldwide (Serpente, 2013). The suit-
ability of the images contained in MBoC to be effectively peda-
gogical was constantly challenged with the latest experimental
outputs and by the constant discussions Roberts and the other
authors had with colleagues regarding the latest results on cell
functioning deriving from experiments of different nature. Exper-
iments from those labs would not only provide the data, but would
set up the boundaries for what an image could visually express and
teach.
For this reason, I suggest, MBoC challenged two taken for
granted conceptions on textbooks. Firstly, that textbooks are pas-
sive carriers of knowledge-a conception that conceptualises peda-
gogy as divorced from research (Garcia-Belmar, Bertomeu-Sanchez,
& Bensaude-Vincent, 2005, pp. 232), and secondly that textbooks
are exemplars of established knowledge as opposed to the fron-
tier science, which is supposedly found in scientic articles only.
I have also argued that molecular images act as pedagogical
tools, because of their capacity to elicit visual based argumenta-
tion, a kind of argumentation which is achieved almost indepen-
dently from textual assertions, a feature that in turn proves
essential for the design of further experiments in current cell
studies. So, although molecular images, because of their deferred
validity, do not act as well-dened devices for settling claims
(Cambrosio, Jacobi, & Keating, 2008), a hint at their making reveals
how important they are because of their capacity to act as both
phenomena and data in the intertwined process of production and
transmission of knowledge. Besides, I contended that the peda-
gogical worth of images is embodied in the process of their
epistemic justication. Both microscopic and molecular images
derive part of their pedagogical power from their capacity to bypass
other forms of argumentation such as propositional reasoning
(Cambrosio et al., 2008, pp. 131). However, molecular images (in
contrast to microscopic images) achieve this from their arrange-
ment in a sequence that mimics experimental operations, from
their integration with other experimental practices and above all
from their re-conceptualisation as data (Cambrosio et al., 2008, pp.
147e148). This is what we saw in the case of Goodsells and Roberts
textbooks imagery, where in addition to all that, the pedagogical
status of molecular images is inseparable from artistic leeway and
the production of scientic data. In short, nowadays perhaps with
more intensity than in the past educating the eye is also justifying
images.
In closing, a pertinent question arises and that is what kind of
archival sources would benet historians interested in dissecting
particular epistemic aspects unfolding in scientic disciplines, such
as the one treated here for the case of cell studies on the rela-
tionship between imagery production, scientic justication and
pedagogy. I am personally attracted to the use of oral history pro-
jects alongside the creation of archives where the whole process of
image production could be captured. This practice should allow
reconstructing the visual intelligence (Jordanova, 2012, pp. 70) at
play during the construction of epistemic claims in times when the
design of visual expressions hinges on a complex interplay between
the authors choices and the combined use of many sophisticated
visual technologies. A visual history based on the sketches pre-
ceding the nal image to feature in textbooks alongside scientists
accounts of the pictorial process would be not only essential to
capture the many concealed craft like aspects of scientic prac-
tices, but to bring closer the work of archivists and historians of
science.
Acknowledgements
I would like to thank Professors Keith Roberts, and Martin Raff
co-authors of MBoC, Drs Jean Paul Vincent and Maximiliano
Rodriguez from The Francis Crick Institute, Mill Hill Laboratory, as
well as to the many former colleagues from the former National
Institute for Medical Research at Mill Hill, with whom I discussed
issues of image justication in the past. I would also like to thank
Professors Keith Roberts and David Goodsell for providing me with
the images used in this paper.
I am grateful to the Department of Science and Technology
Studies of University College London in particular to his head
professor Joe Cain for his support.
Finally, special thanks to Drs Christine Aicardi and Miguel
Garca-Sancho for inviting me to the one-day conference at the
origins of this special issue held in 2013 at the Wellcome Trust,
London and above all for helping me to made of this an intelligible
manuscript.
References
Aicardi, C. (2016). Francis Crick, cross inuencer: A narrative model to historicize
big bioscience. Studies in History and Philosophy of Biological and Biomedical
Sciences, 55, 83e95.
Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., & Watson, J. D. (1983/1989/1994).
Molecular biology of the cell. New York, London: Garland Publishing, Inc.
Alberts, B., Johnson, A., Lewis, J., Morgan, D., Raff, M., Roberts, K., et al. (2014).
Molecular biology of the cell. New York: London: Garland Publishing, Inc.
Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K., & Walter, P. (2002/2008).
Molecular biology of the cell. New York: London: Garland Publishing, Inc.
Alston, W. P. (1989). Epistemic justication: Essays in the theory of knowledge. Ithaca
and London: Cornell University Press.
Anderson, N., & Dietrich, M. R. (Eds.). (2012). The educated eye: Visual culture and
pedagogy in the life sciences. Hanover: Dartmouth College Press.
Ayer, A. J. (1956). The problem of knowledge. Middlesex: Penguin Books.
 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
Baigrie, B. S. (1996). Picturing Knowledge: Historical and philosophical problems
concerning the use of art in science. Toronto, Buffalo, London: University of
Toronto Press.
Bauer, H. (1986). Frontier science and textbook science. Science & Technology
Studies, 4, 33e34.
Bauer, H. (1992). Scientic literacy and the myth of the scientic method. Urbana &
Chicago: University of Illinois Press.
Bechtel, W. (2006). Discovering cell mechanisms: The creation of modern cell biology.
Cambridge University Press. Cambridge. New York. Melbourne. Madrid. Cape
Town. Singapore, Sao Paulo.
Bogen, J., & Woodward, J. (1988). Saving the phenomena. Philosophical Review, 97,
303e352.
Bonjour, L., & Sosa, E. (2003). Epistemic justication: Internalism vs, externalism,
foundations vs. virtues. Blackwell Publishing Ltd. Malden USA, Oxford UK, Vic-
toria Australia, Berlin. Germany.
Brown, J. R. (1996). Illustrations and inferences. In B. S. Baigrie (Ed.), Picturing
knowledge: Historical and philosophical problems concerning the use of art in
science (pp. 250e268). Toronto, Buffalo, London: University of Toronto
Press.
Cambrosio, A., Jacobi, D., & Keating, P. (1993). Ehrlichs Beautiful Pictures and the
controversial beginnings of immunological imagery. Isis, 84, 662e699.
Cambrosio, A., Jacobi, D., & Keating, P. (2006). Arguing with images: Pauling theories
of antibody formation. In L. Pauwels (Ed.), Visual cultures of science: Rethinking
representational practices in knowledge building and science communication (pp.
153e193). Hannover: London University Press of New England.
Cambrosio, A., Jacobi, D., & Keating, P. (2008). Phages, antibodies and de-Mon-
stration. History & Philosophy of Life Sciences, 30, 131e158.
de Chadarevian, S. (2016). The future historian: Reection on the archives of
contemporary sciences. Studies in History and Philosophy of Biological and
Biomedical Sciences, 55, 54e60.
Daston, L. (2014). Beyond representation. In C. Coopmans, J. Vertesi, M. Lynch, &
S. Woolgar (Eds.), Representation in scientic practice revisited (pp. 319e322).
Cambridge, Massachusetts. London, England: The MIT Press.
Daston, L., & Lunbeck, E. (Eds.). (2011). Histories of scientic observation. Chicago &
London: The University of Chicago Press.
De Chadarevian, S., & Kamminga, H. (Eds.). (1998). Molecularizing biology and
medicine: New practices and alliances, 1910se1970s. Amsterdam: Harwood Ac-
ademic Publishers.
Elkins, J. (1995). Art history and images that are not art. Art Bulletin, LXXVII(4), 553e
571.
Elkins, J. (1999). The domain of images. Ithaca & London: Cornell University Press.
Fangerau, H., Chhem, R., Muller, I., & Wang, S. C. (Eds.). (2012). Medical imaging
and philosophy: Challenges reections and actions. Stuttgart: Franz Steiner
Verlag.
Fleck, L. (1979). Genesis and development of a scientic fact. original in German, 1937.
Chicago & London: The University of Chicago Press.
Gail, F. (2003). Plato on knowledge and forms: Selected essays. Oxford: Oxford Uni-
versity Press.
Galison, P., & Jones, C. A. (Eds.). (1998). Picturing science: Producing art. New York:
Routledge.
Garcia-Belmar, A., Bertomeu-Sanchez, J. R., & Bensaude-Vincent, B. (2005). The
power of didactic writings: French chemistry textbooks of the nineteenth
century. In D. Kaiser (Ed.), Pedagogy and the practice of science: Historical and
contemporary perspectives (pp. 219e252). Cambridge, Massachusetts. London:
England.
Garcia-Sancho, M. (2016). The proactive historian: Methodological opportunities
around the new archival evidence on modern genomics. Studies in History and
Philosophy of Biological and Biomedical Sciences, 55, 70e82.
Gaster, B. (1990). Assimilation of scientic change: The introduction of molecular
genetics into biology textbooks. Social Studies of Science, 20, 431e454.
Gettier, E. (1963). Is justied true belief knowledge? Analysis, 23, 121e123.
Giere, R. N. (1996). Visual models and scientic judgements. In B. S. Baigrie (Ed.),
Picturing knowledge: Historical and philosophical problems concerning the use of
art in science (pp. 269e302). Toronto, Buffalo, London: University of Toronto
Press.
Godfrey-Smith, P. (2003). Theory and reality: An introduction to the philosophy of
science. Chicago and London: The University of Chicago Press.
Gomperts, B. D., Kramer, I. M., & Tatham, P. E. R. (2002). Signal transduction.
Amsterdam, London: Elsevier Academic Press.
Gooding, D. C. (1990). Magnetic curves and the magnetic eld: Experimentation
and representation in the history of a theory. In D. C. Gooding, T. Pinch, &
S. Schaffer (Eds.), The uses of experiment: Studies in the natural sciences (pp. 183e
224). Cambridge: Cambridge University Press.
Gooding, D. C. (2004). Envisioning explanations- the art in science. Interdisciplinary
Science Reviews, 29, 278e294.
Goodsell, D. S. (1991). Inside a living cell. Trends in Biochemical Sciences, 16, 203e
206.
Goodsell, D. S. (1992). A look inside the living cell. The American Scientist, 80, 457e
465.
Goodsell, D. S. (1983/2009). The machinery of Life. Berlin, Heidelberg, New York:
Springer-Verlag.
Goodsell, D. S. (2005). Visual methods from atoms to cells. Structure, 13, 347e354.
Goodsell, D. S., & Johnson, G. (2007). Filling in the gaps: Artistic license in education
and outreach. PLoS Biology, 5, 2759e2762 Accessed 10.02.13 http://www.
plosbiology.org.
115
Hacking, I. (1985). Do we see through a microscope? In P. M. Churchland, &
C. A. Hooker (Eds.), Images of science: Essays on realism and empiricism Chicago,
London: The University of Chicago Press.
Hanson, N. R. (1960). Is there a logic of scientic discovery. The Australian Journal of
Philosophy, 38, 91e106.
Heldin, C.-H., & Purton, M. (1996). Signal transduction. London, New York, Tokyo,
Melbourne., Madras: Chapman & Hall.
Hillman, H., & Sartori, P. (1980). The living cell. Chichester: Packard Publishing Ltd.
Hoyningen-Huene, P. (1987). Context of discovery and context of justication.
Studies of History and Philosophy of Science, 501e515.
Hoyningen-Huene, P. (1993). Reconstructing scientic revolutions: Thomas S. Kuhns
philosophy of science. London. Chicago: The University of Chicago Press.
Hoyningen-Huene, P. (2006). Context of discovery versus context of justication
and Thomas Kuhn. In J. Schickore, & F. Steinle (Eds.), Revisiting discovery and
dustication: Historical and philosophical perspectives on the context distinction
(pp. 119e132). Dordrecht: Springer.
Hughes, A. (1959). A history of cytology. London. New York: Abelard-Schuman.
Iversen, M. (1993). Alois Riegel art history and theory. Cambridge, Massachusetts:
MIT Press.
Jordanova, L. (2012). The look of the past: Visual and material evidence in historical
practice. Cambridge, UK: Cambridge University Press.
Kaiser, D. (2005). Pedagogy and the practice of science: Historical and contemporary
perspectives. Cambridge, Massachusetts. London, England: The MIT Press.
Kay, L. (1993). The molecular vision of life: Caltech, the Rockefeller foundation and the
rise of the new biology. New York, Oxford: Oxford University Press.
Kemp, M. (2000). Visualizations: The nature book of art and science. Berkeley. Los
Angeles: University of California Press.
Kemp, M. (2006). Seen/unseen: Art, science, and intuition from Leonardo to the hubble
telescope. Oxford. New York: Oxford University Press.
Kemp, M. (2014). A question of trust: Old issues and new technologies. In
C. Coopmans, J. Vertesi, M. Lynch, & S. Woolgar (Eds.), Representation in scientic
practice revisited (pp. 343e346). Cambridge, Massachusetts. London, England:
The MIT Press.
Kendrew, J. C. (1966). The thread of life: An introduction to molecular biology. Har-
vard: Harvard University Press.
Kepes, G. (1956). The new landscape in art and science. Chicago: Paul Theobald and
Co.
Knorr-Cetna, K. D., & Amann. (1990). Image dissection in natural scientic enquiry.
Science Technology and Human Values, 15, 259e283.
Kordig, C. (1978). Discovery and justication. Philosophy of Science, 45, 110e117.
Kuhn, T. S. (1970). The structure of scientic revolutions. Chicago. London: The Uni-
versity of Chicago Press.
Kuhn, T. S. (1977). The essential tension: Selected studies in scientic tradition and
change. Chicago and London: The University of Chicago Press.
Landesman, C. (2002). Scepticism: The central issues. Oxford UK, Malden USA:
Blackwell Publishers Ltd.
Latour, B., & Woolgar, S. (1986). Laboratory life: The construction of scientic facts.
Princeton: Princeton University Press.
Lehninger, A. L. (1973). A short course in biochemistry. New York: Worth.
Lehrer, K. (2000). Theory of knowledge. Colorado, Oxford: Westview Press.
Lynch, M. (1990). The externalised retina: Selection and mathematization in the
visual documentation of objects in the life sciences. In M. Lynch, & S. Woolgar
(Eds.), Representation in scientic practice (pp. 153e186). Cambridge, Massa-
chusetts. London. England: The MIT Press.
Lynch, M. (2014). Representation in formation. In C. Coopmans, J. Vertesi, M. Lynch,
& S. Woolgar (Eds.), Representation in scientic practice revisited (pp. 323e328).
Cambridge, Massachusetts. London. England: The MIT Press.
Manghani, S., Piper, A., & Simon, J. (Eds.). (2006). Images: A reader. Los Angeles,
London, New Delhi, Singapore. Washington DC: SAGE Publications Ltd.
Miller, M. K. (2002). Drawing conclusions: Art and science combine to image what
cannot be seen. An interview with David Goodsell. http://www.exploratorium.
edu/traits/goodsell.html. (Accessed 20.04.12).
Moser, P. K. (1996). Empirical knowledge: Readings in contemporary epistemology.
Maryland, London: Rowman & Littleeld Publishers Inc.
Musgrave, A. (1993). Common sense science and scepticism: An historical introduction
to the theory of knowledge. Cambridge: Cambridge University Press.
Neta, R., & Pritchard, D. (2009). Arguing about knowledge. London & New York:
Routledge.
Nickles, T. (1980). Scientic discovery and the future of philosophy of science. In
T. Nickles (Ed.), Scientic discovery logic and rationality (pp. 1e60). Dordrecht.
Boston, London: D Reidel Publishing Company.
Olson, A. J., & Goodsell, D. S. (1992). Visualizing biological molecules. Scientic
American, 267, 76e81.
OHara, R. J. (1996). Representations of the natural system in the nineteenth cen-
tury. In B. S. Baigrie (Ed.), Picturing knowledge: Historical and philosophical
problems concerning the use of art in science (pp. 164e183). Toronto, Buffalo,
London: University of Toronto Press.
Pauwels, L. (2006). Visual cultures of science: Rethinking representational practices in
knowledge building and science communication. Hanover. London: Dartmouth
College Press.
Perini, L. (2010). Scientic representation and the semiotics of pictures. In
P. D. Magnus&, & J. Busch (Eds.), New waves in philosophy of science (pp. 131e
154). New York: Macmillan.
Pfeiffer, L. (2014). Encyclopaedia Britannica Accessed 15.08.14 http://www.
britannica.com/EBchecked/topic/1390162/Fantastic-Voyage.
116 N. Serpente / Studies in History and Philosophy of Biological and Biomedical Sciences 55 (2016) 105e116
Popper, P. (1959). Logic of scientic discovery. New York: Harper & Row.
Rasmussen, N. (1993). Facts, artifacts, and mesosomes: Practising epistemology with
the electron microscope. Studies on History and Philosophy of Science, 24, 227e265.
Rasmussen, N. (1995). Mitochondrial structure and the practice of cell biology in
the 1950s. Journal on the History of Biology, 28, 281e429.
Rasmussen, N. (1997). Picture control: The Electron microscope and the transformation
of biology in America, 1940e1960. Stanford: Stanford University Press.
Reichenbach, H. (1938). Experience and prediction. Chicago: University of Chicago Press.
Reisch, G. A. (2005). How the cold war transformed philosophy of science: The icy
slopes of logic. , Cambridge: Cambridge University Press.
Rheinberger, H.-J. (1997). Towards a history of epistemic things: Synthetizing proteins
in the test tube. Standford, California: Standford University Press.
Richardson, A. (2007). That sort of everyday image of logical positivism: Thomas
Kuhn and the decline of logical empiricist philosophy of science. In
A. Richardson, & T. Uebel (Eds.), The Cambridge companion to logical empiricism
(pp. 346e370). Cambridge: Cambridge University Press.
Richter, M. I. (2014). Scientic visualisation: Epistemic weight and surpluses. Frank-
furt: Peter Lang GmBH.
Roberts, K. (2003). On drawing molecules. In J. Inglis, J. Sambrook, & J. A. Witkowski
(Eds.), Inspiring science: Jim Watson and the age of DNA (pp. 435e442). Cold
Spring Harbour New York: Cold Spring Harbour Laboratory.
Robin, H. (1992). The scientic image: From cave to computer. New York: W H
Freeman and Company Publishers.
Schaffner, K. (1974). Logic of discovery and justication in regulatory genetics.
Studies in History and Philosophy of Science, 4, 349e385.
Schefer, I. (1965). Conditions of knowledge: An introduction to epistemology and
education. Glenview. Illinois: Scott, Foresman and Company.
Schickore, J., & Steinle, F. (Eds.). (2006). Revisiting discovery and justication: His-
torical and philosophical perspectives in the context distinction. Dordrecht. The
Netherlands: Springer.
Serpente, N. (2011a). Cells from icons to symbols: Molecularising cell biology in the
1980s. University of London. PhD thesis.
Serpente, N. (2011b). Cells from Icons to symbols: Molecularising cell biology in the
1980s. Studies in History and Philosophy of Biological and Biomedical Science, 42,
403e411.
Serpente, N. (2013). Beyond a pedagogical tool: The 30th anniversary of the pub-
lication of Molecular Biology of the Cell (1983e2013). Nature Reviews Cell and
Molecular Biology, 14, 120e125.
Shaw, J. (2016). Documenting genomics: Applying archival theory to preserving the
records of the Human Genome Project. Studies in History and Philosophy of
Biological and Biomedical Sciences, 55, 61e69.
Smith, B. K., & Blankinship, E. (2000). Justifying imagery: Multimedia support for
learning through explanation. IBM System Journal, 39, 749e767.
Soon Park, B. (2005). In the Context of Pedagogy: Teaching strategy and theory
change in quantum chemistry. In D. Kaiser (Ed.), Pedagogy and the practice of
science: Historical and contemporary perspectives (pp. 287e322). Cambridge,
Massachusetts. London, England: The MIT Press.
Warwick, A. (2003). Masters of theory: Cambridge and the rise of mathematical
physics. Chicago: University of Chicago Press.
Watson, J. D. (1965). Molecular biology of the gene. New York, Amsterdam: W. A.
Benjamin.
Whewell, W. (1847). The philosophy of the inductive sciences, founded upon their
history. London: John W Parker.
Woodward, J. (1989). Data and phenomena. Synthese, 79, 393e472.
Zampieri, F., Zanatta, A., & Bonati, M. R. (2012). Iconography and wax models in
Italian early smallpox vaccination. In H. Fangerau, R. Chhem, I. Muller, &
S. C. Wang (Eds.), Medical imaging and philosophy: Challenges reections and
actions. Stuttgart: Franz Steiner Verlag.