Images are used in science to present data. Taking the physical and life sciences as paradigms, one might suppose that machine-made images always serve this purposes better than hand-made images. A broader view of the use of scientific images, which includes the social sciences, shows this supposition to be false. Archaeological drawings of human artifacts, such as stone tools, are preferred to photographs. The explanation of this is that the human drawing system is an observational tool with special powers.

A teeming variety of images serves a variety of scientific purposes. Scientific images include geometrical diagrams, Doppler radar images, bubble chamber photographs, fMRI scans, Feynman diagrams, and maps of adaptive landscapes. These and other images are used to formulate hypotheses, conduct proofs, describe experimental set ups, confirm hypotheses, present and analyze data, test and calibrate instruments, and communicate data, results, or hypotheses. One might expect that in any given scientific context, some kinds of images will serve some purposes significantly better than other kinds of images. Thus a full taxonomy of scientific images will reveal systematic connections between (1) image types, (2) purposes, and (3) contexts of use. However, philosophers of science (unlike historians of science) have focussed on two small corners of the ground such a taxonomy should cover – diagrams used in mathematical reasoning, on one hand, and, on the other, machine-made images used to present data in the physical and life sciences. A look at images falling outside these areas of focus is needed for progress towards the full taxonomy; it also affords an opportunity to reexamine some standard historical and philosophical accounts of scientific images.

Lithic Drawing

Archaeological illustrations fall into two main classes. Receiving greater attention are 'reconstruction drawings,' which depict imagined scenes of prehistoric hunting, foraging, campfire life, childrearing, and other such activities. According to Stephanie Moser, these drawings were used by archaeologists to argue in a distinctively visual manner for or against conferring hominid status on fossil specimens of australopithecenes and other primates (Moser 1996, 1998). While the drawings neither constitute nor present evidence for or against hominid status, they seem to have been persuasive, and so make for a rich case study of images in scientific rhetoric. Receiving less attention are images of archaeological finds. Settlements and other sites are mapped and buildings are rendered in architectural plan view. Surprisingly, though, artifacts (such as tools) are almost invariably drawn – photographs appear only in popular, non-professional books (e.g. Waechter 1976). Why are drawings preferred? What roles do they fill in archaeological practice? These questions might have different answers for different types of artifact, so consider the case of flaked stone tools.

The authors of a recent handbook written for archaeologists and illustrators feel obliged to explain why drawing is preferable to photography in archaeology. They write that whereas photography "has the disadvantage of being unselective," "a drawing can convey much more relevant and comparable information and can be edited more easily than a photograph" (Adkins and Adkins 1989: 6). Adkins and Adkins give two explanations of the importance of selectivity in representing artifacts. First, "a good drawing selectively portrays the details that the reader needs to see and edits out irrelevant details, so that the illustration can be understood much more easily" (Adkins and Adkins 1989: 7). Second, "the purpose of the illustration is to convey not only information but also an interpretation of that information" (Adkins and Adkins 1989: 5). The first explanation is in tension with the fact that photographs are common in popular publications, where easy understand would seem to be paramount. The second explanation raises questions about the role and nature of interpretation in archaeology.

Figure 1

To find out what an artifact is for, examine the details of its making: this principle works just as well for artifacts that are drawings of artifacts as it does for artifacts that are drawn.

Luckily, the details of lithic drawing are lucidly and systematically set forth in Lucile Addington's manual for archaeological illustrators (Addington 1986). Flaked stone tools are drawn either by an archaeologist or, more often, by a collaborator with specialized training in lithic drawing. Addington is such an illustrator, and her manual is intended to be read by illustrators-in-training and also by archaeologists seeking explicit instruction on the principles of lithic drawing.

As the manual makes clear, lithic drawings must serve two needs. First, a "universal language" is needed to ensure that drawings from any lab can be compared easily and reliably (Addington 1986: ix, xiii-xiv). Note that photography answers this need perfectly well. Second, the drawings should convey all and only relevant information about the objects represented. What is relevant is determined by the nature of the object represented and the role of the representation in its context of use. Lithic drawings represent artifacts and they represent them as artifacts – as objects whose form is largely the result of deliberate human activity – in a context where they are objects of study for their artifactual characteristics. As Addington remarks, "properly drawn artifacts are invariably more informative than photographs in illustrating a prehistoric knapper's workmanship as well as an artifact's form and diagnostic features" (1986: ix).

Knapping involves repeatedly striking a stone, usually flint, in order to remove flakes. Either the flakes themselves are knapped further or the stone from which they are taken is knapped to leave behind a useful pattern of flake scars. The stone is thus 'retouched' in order to achieve a finished form, and it may subsequently be subjected to reuse or weathering that changes its shape or surface. Flaking stone is not a way to pass the time. Stone tools were made for use as knives, cleavers, scrapers, arrowheads, and hammers. Thus a lithic drawing must show how a stone was knapped and with what purpose in view. To do this, it must show: scale; the pattern, sequence, direction, and force of blows to the stone; the bulb and platform of percussion; areas of retouch, snapping, and truncation; areas of grinding, battering, or abrasion; fractures caused by heating; the effects of materials; and pitting and sickle sheen. Fossils, variegated coloration, patina, seams, banding, and crystalization are not shown.

Since relevant features of each stone must be shown in a way that allows for reliable comparisons, lithic drawing conforms to several conventions. The object is illuminated from 45° in the upper left. The ventral surface of the stone is placed at the bottom of the drawing and, when multiple views are shown, profile views are shown next to the edges they display. The scale is fixed at 1:1. Paper types, drawing instruments, and techniques of pencilling and inking are standardized. Finally, a well-defined representational vocabulary is strictly followed. Stippling indicates the cortex of the stone, with greater density indicating greater roughness. Outlines show flake scars in their sequence of making. Lines imitating the ripples caused by flaking indicate the direction of a blow and, by their thickness, its approximate force. They are shown as attached to only one side of a flake scar and to two sides of a bulbar surface. Curved direction lines indicate snapped surfaces and thermal fractures are shown by spider lines, spoked lines, or swirls. Moreover, stippling and direction lines are dual purpose, since they convey the volume of the stone as well as its surface details. This vocabulary of marks helps replicate the look of the object. It is supplemented by a non-mimetic vocabulary. Arrows point to bulbs of percussion. Dashed lines show where broken fragments of a stone fit together. Tick marks coordinate multiple views so that key features of the stone can be matched up.

Figure 2

In order to make a drawing complying with these rules, an illustrator must learn to 'interpret' or "read an artifact's surfaces" (Addington 1986: 2). He or she must have an eye for flake scars and what they indicate about their making; for the different types of flaking, including retouch, snapping, truncation, grinding, and cleaving; for the effects of materials (flint is most common but other types of stone are also knapped); for the effects of weathering and reuse, which must be distinguished from the effects of knapping; and finally for the practical challenges facing a knapper who is knapping a particular stone, since every stone is different and not all 'readings' of a stone are consistent with what a knapper can do with it. An archaeologist who is not also an illustrator is capable of the same reading, but only an illustrator is able to embody the reading in a drawing.

Drawing in History and Philosophy

Lithic drawings challenge several generalizations that have been made about scientific images and thereby raise some interesting philosophical questions.

According to one standard history (see Topper 1996), early scientific images are for the most part drawings that represent types – botanical and anatomical drawings are prime examples. In many cases, what is represented has features never found in a particular specimen – it is an idealization. In other cases, the object represented has the features of a particular specimen, but it is understood that we are to abstract from its particularity and see it is as standing for a type. Later, with the invention of mechanical imaging, many scientific images are made that represent particulars and not types: we are not to abstract from their particularized features since they present evidence about the particulars themselves. They fix observations for later examination.

Exceptions to this history are hand-drawn images used in mathematical proofs and the presentation of theorems or hypotheses, from Euclid to Feynman. However, these exceptions confirm the explanation of the history: drawings represent types; the representation of particulars comes along later, with mechanical imaging. Thus drawings still have a place in the age of mechanical image-making because they represent types. David Topper writes that "despite the invention of photography about a century-and-a-half ago… the artist still has a role to play in illustration, for the camera captures an individual specimen (the particular) whereas an artist can depict the archetype" (1996: 234).

There are other exceptions. Galileo's drawings of the moon, which were made during the early period, represent a particular. The discovery that the moon's dark spots are shadows caused by mountains required meticulous attention to the specific features of the spots. In the later period, Wegener used drawings of particulars in his accounts of continental drift. Again, the exact details of the shape of continental coastlines was needed in order to establish that the continents were once parts of a single land mass. Nevertheless, the main idea stands. Had they been available, photographs would have worked at least as well and probably better for the purposes of Galileo and Wegener.

Another element of the history is worth mentioning. It is natural to consider scientific images as art, when they are drawings, especially when they are drawings from life and most especially when they are old. At least, it is natural in these cases to put in play the questions of whether they are art and thus of the links between art and science. So one might think of the history as gradually putting a distance between artistic images and scientific images. The point is not that scientific images come to lack aesthetic value – on the contrary, they often have great aesthetic value. Rather, the point is that any aesthetic value they have is merely a side effect of their primary function.

Twinned with the standard history is a standard epistemology. The epistemology has two forks, both concerning what counts as good evidence for empirical generalization. The first fork is the claim that good evidence for an empirical generalization should issue from a reliable process. Add the assumption that processes are the more reliable as they rely on mechanical means of taking measurements, and we have a good epistemic reason to prefer photographs to drawings when we wish to capture data for empirical generalization. The second fork is the claim that good evidence for empirical generalization should issue from an objective process. Add the assumption that data-imaging processes are objective only if they produce images with non-conceptual content. An image has non-conceptual content just in case its having the content it has does not require that whoever or whatever makes the image possess concepts of the properties represented (Lopes 1996: ch. 9). If photographs, unlike drawings, have non-conceptual content, then we have another good epistemic reason to use photographs rather than drawings to capture data for empirical generalization. Note that the forks are linked if objectivity promotes reliability.

This epistemology bolsters the thesis that drawings largely give way to machine-made images in science as images come to be used to capture data for empirical generalization. Data capture must be reliable and objective; machines have both virtues; draftsmen have neither. It also explains the observation that machine-made scientific images and artistic drawings follow diverging paths: the epistemic strictures on scientific image-making trump the aesthetic concerns that give the rule to art.

Lithic illustrations are drawings, but they do not represent types; instead, they capture data about particulars which are used in making empirical generalizations. While the drawings are often elegant, visually interesting, even beautiful, that is incidental to their primary purpose. Moreover, lithic drawings are relatively new, postdating the invention of inexpensive and reliable photography: there is a real preference for them over machine-made alternatives. In these ways, they are exceptions to the standard history of scientific images and the explanation of that history. And they challenge the standard epistemology, too. They are made by hand, which seems relatively unreliable, and their contents depend on the conceptual resources of the illustrators who make them, so they would seem not to be objective. If they are reliable and objective, taking into account the purposes for which they are made, then some account is needed of how that can be so.

Drawing, Reliability, and Objectivity

One might say that lithic drawings fail to capture good data for empirical generalization. Assume that this is not the case. Making this assumption has a price, of course. We must make sense of the epistemic quality of lithic drawings.

Images belong to systems. Each system is individuated by rules that set out how images in the system are made, by marking surfaces mechanically, through digital processing, or by hand-drawing, such that selected determinable features of objects or scenes are represented. Some systems are hierarchically nested within others. For instance, colour photography is nested within black and white photography, since colour photography is black and white photography with the addition of resources for depicting hue and saturation. Stopping at this point, of course, leaves us with a very coarse-grained conception of imaging systems, and a great deal more could be said – and has been said (e.g. Lopes 1996, Kulvicki forthcoming). But this is enough for now.

An imaging process outputs images belonging to a system. It is reliable in representing some determinable property of the world just in case the contents of images in the system are counterfactually dependent on the determinates of the determinable. An image's content is counterfactually dependent on trajectories just in case it represents objects' trajectories and would also represent the trajectories were they different. An image's content is counterfactually dependent on the outline shape of a stone just in case it represents its outline and would also represent the outline were the stone a different shape. Imaging processes which generate images whose contents are counterfactually dependent in this way are reliable.

One way to ensure that an imaging process is reliable is to build a device that makes images mechanically. Not all imaging devices work reliably, of course, but many do. Film-based photography is highly reliable for many determinables, for example. However, it is a mistake to assume that the only way to ensure counterfactual dependence is through mechanical causation. A system of images made by drawing is reliable in representing, say, shape, so long as the drawings represent an object's shape and would have represented it as having a different shape were its shape different. Do drawing systems meet that condition? Some do.

The conventions and techniques of lithic drawing ensure its reliability with respect to a selection of determinables, such as the scale of a stone, its overall shape, the shape and location of flake scars and retouch, and the location of features like thermal fractures. That is, an illustrator who follows the rules will produce drawings which show the relevant features of stones and which would also have shown those features as different, were they different. This counterfactual dependence is not a fluke; it is a consequence of compliance with the conventions and techniques of lithic drawing; and that is why lithic drawing is a reliable process.

This explanation of why lithic drawing is reliable might appear to preempt its objectivity. After all, in complying with the conventions of lithic drawing, an illustrator must exercise concepts such as 'flake scar,' 'retouch,' and 'thermal fracture.' He is not merely drawing the look of the stone; he is drawing its look given a reading of it in terms of those concepts. (You and I would make poor lithic illustrators, no matter what out drafting abilities, because we lack the required concepts.) However, we have assumed that drawing, to be objective, cannot require that an illustrator possess concepts of the features depicted. So lithic drawing, for the very reason that it is reliable, is not objective.

A perfectly reasonable move at this juncture denies the assumption that image-making processes are objective only if they are nonconceptual. Objectivity is consistent with observations or data being theory-laden, otherwise few if any processes of data capture are objective. Moreover, mechanical image-making processes are selective in what they represent and designing selective mechanisms usually involves the exercise of concepts of the selected determinables. Nobody designs fMRI systems without making use, as part of the design process, of the concept of neural activation. Nevertheless, as reasonable as it may be, this move fails to address what is worrying about lithic drawing. Lithic drawing requires the illustrator to 'read' or interpret each stone as part of rendering it on paper. The rendering therefore presents an interpreted data point, and the trouble is that one cannot establish a good empirical generalization on the basis of individually interpreted data points.

What we need, if lithic drawings are objective, is a notion of objectivity that is consistent both with the penetration of data by theory and also with the fact that lithic drawings issue from individual interpretations of stones.

Consider what is special about lithic drawing. Lithic illustrators must 'read' a stone's surface and then represent their reading using the conventionalized vocabulary of the system. But anyone who is familiar with flint knapping and with the vocabulary of lithic drawing – any archaeologist, for example – can do as much. The illustrator does something more and it is something that archaeologists who are not illustrators cannot do. The illustrator draws the stone. That is to say, the illustrator has an eye and hand that are able to make marks on the paper that capture the finely detailed contours and textures of the stone. These features are not ones of which the illustrator must have concepts. Thus, the content of a lithic drawing is not wholly a product of interpretation.

Something more is needed for an explanation of how lithic drawing is objective in a useful sense. The proposition that the contents of lithic drawings are not wholly products of interpretation does not erase the worry that some contents of lithic drawings are the products of interpretation and that the contents in question are exactly those that are meant to furnish evidence for archaeological hypotheses.

This is a moderate objectivity and it may be worth having. Topper comments that "theory-ladenness in depiction… is a double-edged sword: on the one hand, concepts aid us in seeing what may otherwise be missed; on the other hand, they can also impede us in recognizing something that does not fit the given categories but which may, in fact, be sitting in front of our noses" (1996: 223). Lithic drawings embody interpretations that are very difficult to achieve from inspecting photographs, and this is what makes them so useful in archaeology. At the same time, however, they do not prevent our seeing alternative interpretations. In this, they possess a kind of objectivity – one that may be worth having, in particular, in social scientific inquiry, when interpretation is both ineliminable and positively desirable.

An important issue has been skirted so far. What weight is borne by the fact that the representations of stone tools used in archaeology are images? The answer is not merely that images show what the artifacts look like. That is news to nobody. Rather, the images show what the stones look like and also capture data about their making in a format that is lends itself to empirical generalization. The case of lithic drawing shows that the two functions are not as far apart as might appear. Counterfactual dependence and non-conceptual content are two marks of the mimetic. They are also, in this case, marks of good evidence.

References

Addington, Lucile R. (1986) Lithic Illustration: Drawing Flaked Stone Artifacts for Publication. Chicago: University of Chicago Press.

Adkins, Lesley and Roy A. Adkins (1989) Archaeological Illustration. Cambridge: Cambridge University Press.

Kulvicki, John (forthcoming) On Images. Oxford: Oxford University Press.

Lopes, Dominic McIver (1996) Understanding Pictures. Oxford: Oxford University Press.

Moser, Stephanie (1996) "Visual Representation in Archaeology: Depicting the Missing Link in Human Origins." In Picturing Knowledge, ed. Brian S. Baigrie. Toronto: University of Toronto Press.

——— (1998) Ancestral Images: The Iconography of Human Origins. Ithaca: Cornell University Press.

Piggott, Stuart (1978) Antiquity Depicted: Aspects of Archaeological Illustration. London and New York: Thames and Hudson.

Topper, David (1996) "Towards an Epistemology of Scientific Illustration." In Picturing Knowledge, ed. Brian S. Baigrie. Toronto: University of Toronto Press.

Waechter, John (1976) Man Before History. Oxford: Phaidon.