Diego Velasquez, in his masterwork on the dynamics of seeing and representation, Las Meninas (Reference Velasquez1656),Footnote ¹ paints himself looking both outward and inward at the same time. He, the artist within the painting, has paused in his work and he ostensibly is looking at us, the viewers, implicating us as part of the scene. His gaze betrays an inwardness, however – his eyes are not focused outwards – coding for us the scene he sees as internal, his own mental vision, thus making the picture he is painting both in and of the external world and the internal world. He shows us the stimulus of the perception of something out there, ‘seeing the picture’, and by his implication, the understanding by the mind which we now know happens in the brain.Footnote ² While Velasquez probably had knowledge of technological aids to representation,Footnote ³ he had to have come to his complex understanding of perception and its relation to representation through a close observation and metacognition about his working process and self-awareness about what he was making. The ‘double vision’ of inside/outside correctly alludes to our current understanding of human cognition of the world ‘out there’ as arising from perceptual cues assembled and interpreted by the brain. The painter did not have available either modern biological understanding or modern instrumentation for studying human cognition. That had to wait two to three hundred years.

Why do I begin reflections on reading with this reference to an old painting? As a non-lawyer, and as a non-scientist,Footnote ⁴ I traverse the word/image cultural divide as it comes up in the context of the law and my teaching of law students. There are several parts to this divide. First, there is a strong cultural bias that thinking is accomplished only with words because language is the medium of thought. Las Meninas is an answer to that. The painting cannot be understood in any full sense without reference to ideas that we would consider properly in the realm of thought but which are, nevertheless, expressed purely visually. Second, the necessity of translating pictures into words in order to make their meanings socially available involves a loss of information as some issues are highlighted and others ignored. I mentioned above the look in some painted eyes; there is a whole lot more in the painting and if we were standing in front of it, talking, all the rest of the work would still be part of our perception as we discussed those eyes. The painting would continue to ‘speak’. Understanding this gap between what we are noticing and what is there is critical if we are going to deploy pictures on behalf of communicative goals, which lawyers are increasingly doing. Third, the digital revolution has changed the relationship between words and pictures in ways that we must understand if lawyers are going to navigate in this new media world.

As someone whose own work alternates between writing and visual production, I experience stress, even visual stress, moving back and forth between these discourses. With the stimulus of the colloquium held at the University College London in February 2006,Footnote ⁵ I decided finally to test my strong intuition that there are differences between reading texts made of words and texts that are pictures by looking at the scientific literature on eye movements in reading because ‘eye movement data reflect moment-to-moment cognitive processes in the various tasks examined’ (Rayner, Reference Rayner1998, p. 372). Work has been done on reading both words and pictures. This research sits right on the boundary between what is outside and what is inside in perception of the kinds of texts I am concerned with. And for an artist, it is particularly interesting because the data are often expressed pictorially in a way that makes them visually richer than just charts of numeric data or graphs. So with the caveat that this cannot be a comprehensive survey, I invite the reader to think with me about the following questions: is there a difference in how we read these two kinds of texts, words and pictures? If so, is it important? And if it is important, how might it be important for the law which is so deeply connected to the written word?

Reading has been a very attractive site for cognitive science research – experimenters can give subjects conscious tasksFootnote ⁶ that can be measured and monitored; reading uses the visual system about which we actually know quite a bit of the neurobiology. For this discussion, the single most important fact is that what we see is a construct made out of sensory inputs assembled through neural activity in different parts of the brain.Footnote ⁷ There is no analogue film running in our heads even if it feels that way. Further, because the neural impulses carrying visual information are passed through the amygdala prior to cortical processing, what we see acquires emotional valence before we have any conscious awareness of this information about our feeling. In the broadest context, words and pictures as perceptions both represent just dataflow coming in from outside to be understood by the brain and processed for meaning. As kinds of texts, though, they are very different from each other.

Perceptual psychologist Rudolph Arnheim addresses the problem directly. A picture is a framed (as in set off, set aside, set out) embodied image expressed in visual terms. ‘Images come at any level of abstractness. However, even the most abstract among them must meet one condition. They must be structurally similar (isomorphic) to the pertinent features of the situations for which the thinking shall be valid’ (Arnheim, Reference Arnheim1971, p. 227). Intuitive cognition, according to Arnheim, ‘takes places in a perceptual field of freely interacting forces’ (p. 233).Footnote ⁸ Pictures are like that. Words, when set down, are made up of marks that, compared to pictures, have much less sensual appealFootnote ⁹ with only conventional relationships to the perceptual/conceptual fields to which they refer. They are a vehicle of ‘intellectual thinking [that] strings perceptual concepts in linear succession. Caught in a four-dimensional world of sequence and spatial simultaneity, the mind operates, on the one hand intuitively by apprehending the products of freely interacting field forces; on the other hand, it cuts one-dimensional paths through the spatial landscape intellectually. Intellectual thinking dismantles the simultaneity of spatial structure. It also transforms all linear relations into one-dimensional successions – the sort of event we represent by an arrow’ (p. 246). Are these distinctions of Arnheim’s (which describe two kinds of texts and two kinds of mental operations that cross each other) echoed in the eye movement research on reading?

Our eyes are continually in motion. Setting aside the small oscillations they make, reading involves two relatively deliberate (because arising from cortical activity) acts: saccade and fixation. When eyes fixate, they align the fovea, the area of clearest focus, highest resolution, on some element in the external world, take in information and move to another site to take in more. The saccade is the movement that occurs when the eye is looking for a new point of fixation. This movement both seeks the next focus and gives some processing time for the data already perceived. Seeing is:

These fixations and movements occur in normal visual perception. And they are a necessary part of the visual system:

We do not notice any impairment in our seeing as our eyes move (assuming normal vision); the visual field is uninterrupted by blurring within our visual angle.

The perceptual task in reading either words or pictures involves fixating on some part of the object of perception and moving our fixations around the visual field of the text under examination. If it is a text made up of words, we get a sense of the form of the whole (whether phrase, sentence, entire work) only when we have completed our reading. If the text is a picture, we can grasp the form of the whole but only through time do we discover the relationship of the parts to the whole that we perceived at the start. If it is a verbal text, we find ‘space’ (its final form as a work) within the time; if it is a picture, we find time (by expending it) in the space. In the case of reading words, we read from word to word or from groups of words to groups of words, even if we are skimming and jumping around. We must keep our fovea (the part of the retina with the highest resolution perceptually speaking) on the words of the text we are reading and move it with a set of rhythms that help us to decode what we are reading. ‘Saccades are directed to definite places in the peripheral vision, and are not merely sent off for some excursion whose extent is determined by internal factors, regardless of what awaits the eye at the end of the saccade’ (Hochberg, Reference Hochberg, Monty and Senders1976, p. 410). Simply put, the peripheral vision works to orient the eye for the next fixation. ‘When readers make two fixations [think when the eye tries again to get useful information] on a word, in the vast majority of cases, an initial fixation on a word is followed by a rightward movement within the word. By far the most frequent pattern is to fixate near the beginning of a word followed by a fixation near the end of the word’ (Rayner, Reference Rayner1998, p. 387). That rightward movement that we readers of English (and other western languages) know so well constitutes a kind of constraint of the visual field that maximises efficiency as we move forward in the text.

When we look at a picture, ‘reading’ does not seem to take any particular effort at the level of perception because visual texts, pictures, are just there – because of those isomorphic relationships to the subject of representation, they may seem not to need further decoding, although, in fact, that is far from the case. Pictures are seemingly automatically parsed right along with our visual perception of the world, naturalised. But this can lead us to error because if we simply identify the picture with its subject matter, we will miss a lot of the meaning. Pictures do not have a syntax (in spite of our regular colloquial use of terms like ‘the language of art’ or ‘pictorial grammar’) and so they can be entered at any place in the visual field and that sense of constraint in reading verbal text is absent – when we go back for more we don’t have to move right. That said, makers of pictures may go to some lengths to make some elements particularly important visually but there is no way to guarantee a sequence of experience. Pictures are in the first instance spatial constructions and therefore we can see a gestalt at the beginning. They take time to read, time to discover all the relationships between pictorial elements, but they do not act like words in meaningful aggregations which unfold through time in syntactical sequences.

Order matters in reading words. Think of the difference in meaning when we take the opening words of Moby Dick and reorder them. Melville wrote ‘Call me Ishmael. Some years ago – never mind how long precisely. . .’ introducing the storyteller and suggesting a mythic theme – signalled by his statement that precise time is unimportant. Changed to ‘Never mind how, call me, Ishmael – precisely – some years ago. . .’ makes it all about the storyteller, less about the story. He comes across as a bit peevish and we are not sure we are willing to spend a lot of pages in his company.

The orderly unfolding of language, experienced in concert with the movement of our eyes as they move across a text, seems to match the unfolding of thought, especially reasoned thinking, to which we give special cultural status. Law privileges reasoned thinking. This is how we may believe such thinking works in our brains, just as we can so easily believe that there is a picture in there because we can see images when we close our eyes. We dream with images; we can remember with images. For both words and images, perception works in a different way than our experience of it:

Eye movement research itself shows us the complexity of the problem, both in issues of experimental design and in the content of the data displays in the published reports. What do the pictures from eye movement research tell us? Guy T. Buswell’sFootnote ¹² eye movement research in the 1920s and 1930s used lights, mirrors, cameras and movie film to track viewers’ fixations on various kinds of texts. There has, of course, been an enormous change since then in instrumentation for recording and synthesising eye movement data, but those initial studies are still found to be useful and quoted in more recent work. All we see in Buswell’s data reports on reading words are sentences with vertical lines drawn through the line of text, perpendicular to it, almost like marking beats in a musical score. There are small numbers that indicate the length of the fixation expressed in units of 30ths of a second. That figure matches the speed of the film moving through the camera but does not indicate the length of a single fixation. Saccades are measured by their amplitude and can vary in length. Microsaccades, which might analogised to the screen saver on your computer monitor, oscillate back and forth at a rate of about 60 per second and ‘are completely imperceptible under normal circumstances. They serve to refresh the image being cast onto the rod cells and cone cells at the back of the eye. Without microsaccades, staring fixedly at something would cause the vision to cease after a few seconds since rods and cones only respond to a change in luminance’ (Wikipedia). Using the spaces between the fixations and from the size of these intervals we can infer from space an idea of time without even reading the numeric labels, which increases the sense of union between the moving eye and the subject of its gaze.

But there is a gap between this representation and what we know about the cognitive actions happening inside the brain. We cannot infer from these pictures the brain tagging and slotting in elements giving a temporal order to the spatial locations referenced above. The data sets suggest an indexical relationship between mark and fixation so that we can even imagine that we have direct access to the data itself, as if the eyes made the marks which, of course, they did not. This sort of indexical relationship, or seeming to be one, underlies the argumentative power of forensic evidence and eyewitness identifications. These indexical relationships can seem to be unarguable although we know that this is not the case.

One of the most remarkable aspects of these data set pictures is that we can immediately see, if not understand the reasons for, the comparative ease with which readers of differing skills read written texts. Good readers have an evenly spaced ‘stroke’ of fixation and proceed through the text (as indicated by the vertical lines standing for fixation points) with the regular rhythm of a great swimmer – fixation, saccade, fixation, saccade. Fixations are shorter, saccades are longer. (Remember that the time of the saccade permits the brain to process some of the perceptual image.) In contrast, poor readers go back to resample the stimulus so repeatedly that they have little processing time. Their data sets have thickets of vertical lines, a very fast beat, maybe even an anxious one. If we think commonsenically that bright people have more brain power at their disposal, we might imagine that they put more cells to the task. These pictures show something different than just massive computational power: skill in this context translates into low effort, fewer mental resources required to accomplish the task.

Buswell’s research enterprise was to show empirically that reading was a cognitive activity and not just ‘implicit speech’. It is noteworthy that he welcomed into his experimental design a lot of cognitive activity that he did not then account for or measure. The exemplar text used in the experiments (see illustrations in Kolers Reference Kolers, Monty and Senders1976, pp. 376–9) is one that to my reading is loaded with social assumptions:

The caption for the picture reads ‘prose of modest difficulty’. In the early twenty-first century this text is hardly neutral prose whether of modest difficulty or not. I wish I had more demographic information on the experimental subject. What social differences were there between the good readers and the poor ones? How did those differences connect with the text in question? Something is left out of Buswell’s ‘picture’ – he assumes a simplicity in his experimental system that isn’t there based on clear boundaries between the subject, the reader, and the text, the object of perception. In fact, we now know that perception itself can be socially mediated and constructed (Berns, 2005). Legal rhetors will need to appreciate how their materials might be socially constructed in entirely different ways depending upon the social contexts of the finders of fact. Those who view the picture as a simple statement will miss relevant information and be less in control of their messages if not the content of their cases.

Buswell’s data maps for pictures are different from those for reading text. They look like the ‘join-the-dots’ games for children where a figure emerges from a seemingly chaotic field. As in recording his data for reading written words, he places dots for fixations, but with data concerning pictures, he connects the fixations with lines representing saccades and numbers them so that the order of reading the picture can be traced. By numbering the sequence we can follow the reader wandering through the picture and see changes of direction, leaps cross space, and so on. These data sets do not reflect rhythms in the same way as the figures for written texts. Because the data is reflected in both dots and lines, and we like to find order in our perceptions, we tend to make figures out of the spaces enclosed by the lines, object-like structures sitting on the field represented by the reproduction of the picture used in testing.Footnote ¹³ Just as our eyes do not shape letters as we read, we do not echo the picture analogically when we read pictures either. Pictures and text may stimulate eye movements but they do not do so analogically, so differences in movements between the kinds of texts do not emerge from their visual properties in this way.

What Buswell’s research showed is that: (1) there are two patterns of perception of pictures: general scanning and more detailed examination of parts the picture and the duration of saccades increases with looking at the picture; (2) there are wide individual differences between subjects as they look (although age is not a significant factor – adults and children are not different as groups); (3) colour and repetitive patterns do not do not strongly affect patterns of eye movement (colour vision is less a feature of peripheral vision); (4) the directions given ‘prior to looking at a picture have a marked influence upon the character of the perception’ (Buswell, Reference Buswell1935, pp. 142–44, quote p. 144). (Like all rhetors, lawyers will try to prime their audiences to understand evidence in terms of their case theory. Because pictures are very powerful, they use them increasingly. Buswell’s data suggest that lawyers need to be careful, though, to allow enough time for the finders of fact to examine the picture in detail if those details are significant for the case. And even more importantly, the lawyer using pictures needs to remember that what seems obvious to counsel may not seem at all obvious to other viewers and that counsel cannot count on viewers automatically having similar responses. Whether to underscore visual meaning verbally or just to allow it to deliver its messages directly is a matter of technique; there are benefits to clarifying and there are benefits to allowing the picture to carry meaning that might be difficult to say directly for many reasons. This should be a matter of active choice rather than accidental effect.) Buswell found that he could prime the responses of the experimental subjects by giving them tasks, but this shows up as different patterns of dots and lines. We do not know whether this different reading of the picture entailed different verbal interpretations of the picture as well; Buswell does not seem to have studied or at least reported this. Legal rhetors would do well to give their audiences tasks to prime responses rather than to just tell them an answer.

The ‘general scanning’ that typified the first encounter with a picture by his experimental subjects, did not mean that the fovea passes over all the parts of the picture, as if the eye was like the mechanical scanners we use today. When we look at the mapping of the eye movements for pictures, we can see that the eye did not fixate on all the areas of the picture ever. This shows us that peripheral vision is important as well, otherwise we would see blank spaces of missing data in the picture we are looking at. The mobility of the retinal image occasioned by head ‘movements, blinking movements of the eyes, saccades, drift, and tremor during fixation on an element of a stationary object. . .prevent the formation of an empty field’ (Yarbus, Reference Yarbus and Riggs1967). In fact, the brain fills in the missing data.

Yarbus explored further the issue of task and patterns of visual scanning in reading pictures, using an apparatus that attached directly to the eye to keep the retinal image stable. His data sets also record points of fixations and lines representing saccades connecting them recorded on photosensitive paper during the testing session. These recordings look much more art-like, much less formalised in comparison with Buswell’s images. Buswell, using film, must have made transcriptions of the data relating the recorded fixations and saccades to the film frame and to the picture that was the challenge text. In carrying out this task, Buswell made his data look very scientific and geometically neat, measured and ruled. Yarbus’ work 30 years later shows data sets that look as if they were made with pen and ink and with an artist’s hand. The more an observing subject is interested in something in the picture, the fuzzier and denser the marks become as fixations pile up in the same areas of the picture. Rather than join-the-dots pictures, these look like nets and webs with stuff caught in them. The lines themselves seem to have kinetic energy and occupy the whole picture space in a very different way from Buswell’s plots. In fact, they look more like Giacometti drawings than machine prints. When both researchers used pictures that had people in them, whether plotted as fixations and saccades or densities, the faces received much more attention. This should come as no surprise – infants can recognise their own mothers’ faces within hours of birth (Eliot, Reference Eliot1999, p. 219, also Sinha, 2005). We must have a propensity to pay attention to faces.

Of the two researchers, Yarbus went more deeply into the question of how patterns of reading pictures altered by task. Taking a famous genre painting by the Russian artist I. E. Repin (1844–1930), An Unexpected Visitor (1884), Yarbus tested subjects by first having them freely view the reproduction of the painting, and then gave them various perceptual tasks: from estimating the material circumstances of the family to attributing ages to the persons represented, remembering their clothing, guessing what they were doing earlier in the scene and estimating how long the unexpected visitor had been away from the family. These all could have been loaded questions depending upon the circumstances of the experimental subjects because he is asking about socio-economic status and family relationships. Given his historical period, one of great social unrest, he might have been asking about a visitor who was returning from a forced labour camp. Who knows? But when Yarbus writes about his study, he seems strangely naïve about the issue of meaning when he asserts:

‘The meaning. . .’ Only one? Is it so clear what it might be? How can we know what the observer’s thought is from this data? Even when we could see the rapid fixations in the crowding lines of Buswell’s recordings of poor readers, we certainly could not know the thought of the reader. Do some elements of pictures attract because they are inherently more interesting or because they are harder to parse? Do they attract because they are visually interesting or because their content is loaded? We do not know this from the data. Lacking syntax, how would knowing the order of the eye’s inspection tell us about the viewer’s thoughts? Is Yarbus not insisting on a transparency of consciousness that is unwarranted given what we know about the brain and how many different areas of the brain contribute to any assessment of meaning? (Admittedly, cognitive science has progressed a lot since Yarbus published but Arnheim was a contemporary of his for all practical purposes.) Some of his naïve realism about the picture might well be attributed to the effects of his experimental situation. Yarbus was, after all, defining the tasks for his experimental subjects, so when he saw the readouts he could legitimately observe close correlations between task and pattern of reading without going further into meaning. Perhaps he fell into this trap, too, because he had so internalised his experimental methodology that it felt completely naturalised and, being visual, ‘out there’ in the world and less ‘in there’ in his own conceptions. Buswell was interested in the formal properties of the visual texts he used in studying reading of pictures but did not address the issue of content, so both researchers failed to take into consideration the impact of the uptake of content on the uptake of form. Buswell was thinking like a formalist and Yarbus more sociologically, but neither fully ‘saw’ the material they were working with. Likewise, counsel using pictures can fall into this trap of assuming a simple communicative situation and overlook problems of extra content arising from the pictorial nature of visual materials. Before strategy can be deployed using visual materials for either direct evidence or documentation, demonstrative evidence, or for argument, the lawyer must start by seeing what is there, including making the effort to imagine how other viewers might see the same materials.

Do these eye movement studies and their data sets in any way confirm Arnheim’s proposition about intuitive cognition and intellectual cognition (his terms for the visual and the verbal)? I think that, as representations, they do, in fact, show linear rhythms moving across a line of text (Arnheim’s intellectual cognition) and fields of forces explored in complex ways (Arnheim’s intuitive cognition) displayed in an empirical system. Further, studying subjects deliberately reading, means that they are engaged in a process that is consciously undertaken, whatever unconscious elements are also present. The researchers may not have been looking at meaning in terms of the subject matter but that does not mean it was not present for the people who participated as subjects.

When Buswell and Yarbus present composite data from multiple test subjects, we can see in their pictures of this data both a statistical person and individual variations, e.g. some readers wandered further around the picture space; more collect around key features of the picture. Whether we view the statistical person as providing a baseline for understanding the individual depends, in part, on how we conceptualise individual psychology and the role of outliers, people who do not fall in the center of the sample. There is intriguing eye movement work done with a class of individuals who as a group are outliers: autistic persons. Instead of giving tasks to subjects, a group of researchers, building on research about face perception in infants and children, were able to show that a group of autistic individuals in the test group demonstrated abnormal patterns of social visual pursuit (Klin, 2002). The researchers recorded eye movements while the subjects watched short clips from the film ‘Who’s Afraid of Virginia Woolf’, a dramatic production with highly charged social interactions among four characters. Autistic viewers focused on the chin and mouth areas while the control group focused on the eyes/brows of speakers. Just as social context can influence perception so can individual differences.

Social context; individual difference – these are terms that suggest ‘frames of reference’ which certainly are relevant. In the two kinds of texts broadly considered – word texts and visual texts/pictures – visual framing needs first to be understood in terms of a different level of analysis. This issue appears in neither Buswell nor Yarbus. Looking at their data sets, it is as if the viewers are plunged into the visual space without any particular cues to the framing edge of the images they are viewing. This may be a result, in part, of the experimental setting, the constraints of the apparatus on the head and eyes, the constraint of the visual field (was anything beyond the picture itself visible to the subject of the test?), the instructions of the experimenter, all those elements of the conditions of the testing situation that Latour would call ‘subtle and multiple transformations’ in the ‘transfers of information’ (Latour, Reference Latour1999, p. 298). The information gained in these studies is mediated by the experimental system and is part of the data; it is not a neutral player in this game.

The framing edge of the visual representation (and here I make space for sculpture as well as two-dimensional visual text) both marks the boundary between the representation and the real context in which it appears, what is inside the picture and outside it, and it changes all the visual dynamics within. The frame conditions what seems close and what far away, what is large and what is small; it can imply action in the field and whether it feels open or tight, crowded or empty. The frame contributes to the perceptual experience of everything within its field.

Written words, or printed words, are remarkably robust as the conditions of representation change. I could sit and write this text in long hand on a yellow pad and it would look very different. It would also signal things about me that might be interpreted very differently by other people. Yellow pad as ‘legal’, yellow pad as ‘draft’, yellow pad as ‘professional’ but the words themselves, assuming a readable handwriting, would still fall in the same order, still address the sequential flow of thought being represented by me as I write; they would in themselves still essentially add up to the same meaning. I do not mean to diminish the importance of graphic design in signaling all those associated values, but only to observe that the perceptual qualities of words are more stable than the perceptual qualities of pictures. We know from the words themselves where the text stops.

One of the more fascinating studies of reading has to do with reading music (Goolsby, 1994). It was found that musicians read no more than eight notes ahead of their hands and that they often fixated on the blank spaces on the musical score. We experience the silences within the score as meaningful when we hear the music. Reading research has shown that spaces are important to text as well.Footnote ¹⁴ When spaces are inserted into long compound nouns in German (Inhoff et al., Reference Inhoff, Radach and Heller1996) and or when spaces are put between characters in Thai which did not traditionally have spaces between characters, the spaces greatly enhance the speed of reading even when they are not semantically meaningful (Kolsom and Gobet, 1997). In pictures, the equivalent to musical silences lies in the spaces between elements but which are shaped by them; we call them negative spaces. Negative spaces are ‘real’ like negative numbers. They are present and the dialogue between volume and void, shape as full and shape as empty, is crucial to the play of picturing. We use shape cues to distinguish letters, punctuation, etc. but most often the contribution of the empty space between these elements does not contribute to the content in any strong way except in certain art forms.

At the beginning I asked: Is there a difference in how we read these two kinds of texts, words and pictures? Research answers yes. If so, is it important? Yes, because it points to different cognitive functions, different aspects of mind. Building a thought piece by piece through words can facilitate reasoned thinking, clarified thought, can enable us to relate elements of our experience to many different kinds of qualities rather like money allows us to relate extremely disparate things on the basis of value. When we ‘see’ a picture, we can grasp the simultaneous presence of many disparate things within a common frame, harmonised through composition but allowed to retain a visible connection to the things to which they refer. This is the groundwork for the final question: how might it be important for the law which is so deeply connected to the written word?

We regard reading books as an essential skill for navigating effectively in our culture; it represents an achievement. Reading words requires deliberate education. Children do not necessarily learn to read books just because they are lying around, whereas they can tell us about a picture they see. We seem not to have to learn to read pictures, especially photographs (and photography and the optical attitude it represents dominate our exposure to pictures now), and our culture has not counted this as an achievement if we do. Further, because of the long association of pictures being used as a means to communicate with those who cannot read words, there are without a doubt long-term social class biases as well.

We think of reading (and its partner, writing) as the essential literacy that has made advanced culture possible. The earliest artifacts of human representation go back to marked bones from 35,000–20,000 BCE believed to represent numbers of some kind; European cave paintings can be dated to 20,000 BCE. It took more than 15,000 years to arrive at Sumerian writing, 3200–3100 BCE (Ifrah, Reference Ifrah2001, pp. 9–10). Writing, of course, fixed spoken language in a form that made it transmissible across distance and through time – the reader did not have to be present, or even alive, at the time the writer (as speaker) composed the words. Setting aside the glories of literature and private writing, surviving very early written texts were mostly about bookkeeping and victories in war, treaties and trade. Writing’s potential for permanence allowed for the creation of public records of all kinds, which had a number of effects: the growth of bureaucratic means of administration, the possibility of producing historical accounts, uniform communications from leaders to subjects. By separating ‘reader and author. . .[it created] the possibility of greater impersonality and perhaps rationality in human affairs’ (Finnegan, Reference Finnegan1988, p. 18). In short, the fundamentals were present for the development of law.

Peter Goodrich, in working on a psychoanalytical jurisprudence, sees in the history of the development of English law a conscious wall built against images, essentially shutting itself off from pictures:

The suppression of imagery that Goodrich is writing about has a kind of logic to it, forcing people, consciously or not, to become readers and writers. Pictures, to become part of a conversation between people, must be translated into words. Words themselves, are a different kind of communicative currency because they are directly a medium of social exchange. Following Arnheim’s reasoning, words, having conventional and not isomorphic relationships to the things to which they refer, make it possible to stabilise social meanings and therefore the law more easily as well as to create and examine ideas independent of the constraints of perception. The law depends upon the written text for the transmission of its own culture of cases and precedents through time and through the extensive bodies of writing that have developed through the institution of the legal academy. It has been above all the creature of the written word – and the spoken word in legal performance.

As I write, the performance of the law increasingly involves pictures, whether as attachments to legal documents or as evidence and argument presented in negotiations or at trial. Every kind of picture might be in play – from graphic displays of numeric data to photographs of many kinds; flow charts and organisational charts to cartoons; animations to video recreations; virtual reality views to multi-media combinations. Each picture comes with the rhetoric of its genre and materials, and the associations that are attached to those. For instance, in the Soham case in England (the trial of a school caretaker for the murder of two young girls), jurors in England were given a disk containing pictures of the evidence that they could take into their jury room for deliberations; they took a disk that was, through its design conventions, alluding to a video game (Sherwin, Reference Sherwin and Elkins2006). Likewise, participants in the Bloody Sunday Tribunal (Whelan, Reference Whelan2005) have used virtual reality views of the sites of the terrible events so that paths and points of view could be identified through computer-animated reconstructions of the sites. The multi-media disk used in the Michael Skakel prosecution in the U.S. allowed the prosecution’s final argument to draw upon sound recordings, their transcripts, and pictorial files in a manner that was one media step short of being a video (Carney and Feigenson, Reference Carney and Feigenson2004). Finally, courtroom presentation software like Trial Director and Sanction IIFootnote ¹⁵ present differing kinds of evidence in simultaneous display in multiple windows, like our desktop computer monitors – so that it is possible to play clips from a video deposition while showing the written transcript, among many possibilities. Legal storytelling deploying these resources will refer to cinematic conventions or to the narratives of video games, especially if counsel chooses to demonstrate a number of logical paths with the aim of showing what s/he believes to be the true one according to the evidence. The legal rhetor will have to decide what meaning is to be carried by pictures, what by words, and be able to move fluidly between them.

These new ways of presenting evidence and legal argument, of constructing legal narratives, are possible because of the digital revolution that has put into the hands of consumers as well as professionals the tools of digital imaging. It is easy to make digital pictures, to disseminate them – not just still pictures but video and audio files as well. Not just the professionals under stand how this is done, juries and others involved in the legal decision making process understand these tools too. Audiences have changed. What is happening in the courtroom and across negotiating tables is a contextually specific instance of a much larger social envelope where political dialogue is carried out as a battle of mass mediated images as well as a matter of barbs and slogans.

Law is now practised on the screens of our televisions and movies in America, so that law stories can be seen almost 24 hours a day.Footnote ¹⁶ Real lawyers for big and little clients may well wind up on those screens themselves, and the growth of litigation public relations as a profession is testament to the need of lawyers to think about the media frame of their cases as well. Lawyers, like many people, have tended to believe that persuasion involves changing people’s beliefs about a question. Recent work in political science shows how framing used by a communication source constructs and defines a social or political issue for its audience. ‘Messages may change attitudes by adding information to an individual’s stockpile of considerations about the issue (belief change), by making particular considerations temporarily more accessible (priming), or by altering the weight of particular considerations (framing)’ (Nelson, Oxley and Clawson. Reference Nelson, Oxley and Clawson1997, p. 221). Their study showed that framing issues can activate existing beliefs and cognitions without appearing overtly to persuade. The frames created by media can become absolutely naturalised and therefore invisible to most. The media now are dominantly visual and lawyers are in these pictures.

In this changed media environment, the law as an epistemology dominated by the idea of words as carriers of the most important legal meaning must find a way to integrate and accommodate what is happening – the invasion of the visual. It must integrate the need for some new understanding about the cognitive effects of this change. People expect to be able to act on information, to put it together for themselves. They expect to see the evidence presented in a way that they can understand. I have been ranging widely – from a masterwork of painting to research on eye movements as clues to cognitive function; the history of writing, from old media to new; to the law as a discipline of words. It is a truism to say that the law is reactive. It looks to analogy and distinction to understand new cases in terms of old; it depends upon long chains of decision making to bind the world together through its words. This dramatic shift to the visualFootnote ¹⁷ in the law represents a moment of change when it is being forcibly challenged to give up some old habits and acquire some new ones.

Deeply embedded in our western tradition is the metaphor that the person who sees is the person who knows. This is a verbal metaphor (the seeing in this sense is happening inside a person’s head where a viewer can only find traces in facial expression if that); it transfers the act of seeing to insight, to actions of the mind (Lakoff and Johnson, Reference Lakoff and Johnson1999, p. 393) and to thought as language. ‘Thought has the properties of a language. Thought is external and public. The structure of thought is accurately representable as a linear sequence of written symbols of the sort that constitute a written language. Every thought is expressible in language’ (Lakoff and Johnson, Reference Lakoff and Johnson1999, p. 546). Above I discussed the eye movement research carried out by Buswell and Yarbus, both for the insight it provides for those interested in the uptake of pictures, currently of great relevance to legal practitioners, and because it represents also an example of serious scholars being bound by their frames of reference and missing some potentially important questions relevant to their work. To be sure, no one can see everything. At the same time, it is extremely important to try to find the places where framing edges block our views and to ask questions. This is especially true for the law right now. The problems that the law addresses have real world consequences; people’s lives are affected and as a civilisation of the law, our fundamental social order may be at stake. Certainly, it is if citizens cease to find the law a credible means of solving conflict.

Pictures, because of their system of reference, have bodies – they are made of materials or, if made of pixels, they still reference other things we can see. By and large we confront them ‘person to person’ parallel to our eyes. If we think of them as not language and therefore not thinking, we will be blinded to the ways in which they do speak:

To give a sense of scale for the rapidity of this change:

It is not surprising that we are just beginning to understand what is happening.

It is ironic that just as we humans managed to build thinking machines that seem to be the ultimate Cartesian expression of mind without body, these same machines have brought us full circle to confront the return of the repressed. The computers that we have built are now changing the world, not just because we can more easily accomplish tasks that would otherwise be impossible due to the vastness of the computations they require, but because we have built a global cultural space of profound implications, a culture of representation – cyberspace – that is itself governed by the rule of images. ‘The screen is the site of the image, and the logic of the image dominates the semiotic organisation of the screen’ (Kress, Reference Kress2003, p. 65), whether it is showing us words or showing us pictures or playing our music accompanied by visual ‘skins’ of our choosing. Our encounters with both words and pictures are increasingly happening on the screen and we can no longer claim that words are the only way we speak to one another, the only containers for our ideas, the only generators of real thinking. While the practice of law has always used technologies available to it, the discipline of law has not always paid attention to the implications of their use. This neglect can no longer be tolerated because of the extent and depth of the changes involved.

Just as a frame (thought of as the border and not just the decorative edge of a container for a picture) delimits the perimeter of a visual text, on the computer screen we are reading words and pictures, often together sharing frames. Certainly they share the larger frame of the monitor itself. This presents new reading tasks that involve moving back and forth between words and pictures that qualify and amplify each other. These hybrids are rather like graphic novels ‘demanding and rewarding mental flexibility and nervous stamina. Consuming them – toggling for hours between the incommensurable functions of reading and looking – is taxing’ (Schjeldahl, Reference Schjeldahl2005). We know that words and pictures call for different reading behaviours. It has been shown that ‘the richer representations of information in pictures require extended encoding durations relative to the encoding of information from text’ (Underwood et al., p. 180) which means, when faced with a multi-media display, there are conflicting demands on our cognitive skills. Multimodality may have its virtues (Kress, 2001) – there are several channels that allow people to use their favoured learning mode – but if not used carefully, legal rhetors may find their audiences becoming confused or less attentive. We have all had the experience of trying to decide whether to listen to a speaker or to read the verbally packed PowerPoint slides, knowing that a choice has to be made or we will not get anything at all out of the presentation. Further, to take full advantage of the powers of multimedia, legal rhetors are going to have to give up full control over their messages by allowing meanings to be filled in by their audiences. If they welcome the participation of their audience they will involve them more powerfully in the meaning-making process but they will have to embrace showing and not just telling. What this means is that they are going to have to get very smart about what is inside the frame of the visual text they are dealing with. The problem is not about recognising what the picture depicts; it is to see what it says to the lawyer who must then be able to imagine what it might say to people different from him/herself. We must not assume a transparency that just is not there, just as we cannot assume that the picture we have in our heads is the movie of our minds. Cognitive science can help by assisting our view inward which will then affect our view outward.

There is ample evidence that we process unconsciously, associatively, and analogically, as well as consciously solve problems. When we are engaged in thinking about something in a goal-oriented manner, we are bound more closely to the parameters of our window of attention – and therefore apt not to see changes in the field, like missing the card pass in sleight of hand magic tricks. Think of the empirical scientists, Buswell and Yarbus, thinking so clearly about how to study fixations and saccades in reading while ignoring other, possibly crucially relevant, elements in the systems they set up. We can be inattentionally blind, and we can be made that way by objects, people, events in the perceptual field that are compelling. Our perceptions can be primed by previous experience (a time-based phenomenon) and anchored in unconscious ways (a spatial metaphor for connection that is more a matter of being?).

From the serial and parallel information processing that must underlie our vision and cognitive capacities inside our heads to the ways we conceive of the world outside (Shanahan and Baars, 2005), time and space are fundamental categories of our experience that are also the basis of one of the fundamental contrasts between the kinds of texts that I have been discussing. We need the resources of both; we need to be able to move between the freely interacting fields of forces and the conceptual linear strings (to use Arnheim’s words) defining texts that are picture and those made of words. I am suggesting that the need to communicate and to persuade in this new world can represent an opportunity and a gift and not just a burden. I believe that this is so because in becoming sensitive to a cognitive mode where dissimilar things are brought into relationship within a frame without having to win or lose – what happens in pictures – the discipline of the law may find new ways to look both inward and outward, to see more with wisdom. Understanding pictures requires understanding perception; it means having the double vision on the world reflected in Velasquez’s eyes.