ABSTRACT

The illumination of a scene influences how that scene is scanned and how it is depicted. This premise, together with assumptions regarding implications for teaching observational drawing, was the basis for a pilot study in the Nova Scotia College of Art and Design (NSCAD) Drawing Laboratory. The pilot provided evidence that helped refine the question and methods described in this expanded study. As in the pilot, participants worked from common stimuli that were lit in two distinct ways. The participants drew for a predetermined period of time while their hand movements were recorded digitally and the entire process was observed firsthand. Over a period of 7 days, five participants each completed four drawings. The 20 drawings were compared and the recordings analyzed. Digital analysis generated the most informative data in that, while light’s influence on drawing strategies proved to be less significant than anticipated, changes in drawing behavior were observed with implications for teaching and learning.
INTRODUCTION

​Since the 15th century, those who draw and paint from life have understood that representing the effects of light in a manner that is constant is necessary for the creation of coherent and credible representations of space and form. “The vivid impressions of form we receive through certain lighting effects have preoccupied Western artists for a long time. . . . This admiration for light’s descriptive and transformative powers persists unabated today” (Enstice & Peters, 2003, p. 145). As described by Shlain (2007), light is one of the “organizing benefits of perspectivist space” (p. 54), an observation that he attributed to the noted art historian Ernst Gombrich. Consistent with this maxim, observational drawing instruction, whether derived from books or occurring in the classroom, regularly includes exercises where the learner is expected to respond to light as value, as it influences perception and representation of space and form. Goldstein (1999) provided a rationale for teaching about the effects of light when he offered: “Like line and shape, value is a versatile compositional tool. Value can increase attention to a form’s direction, weight, or importance; it creates visual diversity but also serves as a strong unifying agent” (p. 80). ​

In the natural world, our experience of how a scene is lit is based almost exclusively on the effects of sunlight. The world is lit from above and, as Myers (1989) described, “[m]ost artificial light sources are also placed overhead [and] we prefer it that way” (p. 117). Such “preference” may easily be tested by viewing two-dimensional images of seemingly convex or concave hemispheres where the lighting shines consistently from one direction. Lit from above, the hemispheres with the shadow at the bottom appear convex, while those with shadows at the top appear concave. By rotating the page, as if by magic, the appearance of the hemispheres is reversed (Solso, 1994). In essence, the brain is seeking order based upon expectations about lighting. But artists’ interest in lighting often goes beyond the familiar and the comfortable, and subject matter may be lit from anywhere, including from within the scene. A clear example of this may be found in the work of Georges de La Tour, where the presence and effects of light, often hidden and from within the scene, add drama and, therefore, content. 

​Writing in Elements of Psychology, Krech, Crutchfield, and Livson (1969) provided a useful starting point for situating tasks pertaining to light and perception. It was offered that “[c]ertain patterns of light and shadow favor the impression of depth. In general, lighter shades of color or brighter intensities tend to be seen as nearer” (p. 194). Provocatively, and citing Hess (1961), who designed experiments to disturb an organism’s early experience with such patterns, Krech went on to suggest that “learning can upset this rule” (p. 215). And, while not solely intended to upset rules, an instructor’s interest in presenting novel drawing challenges to students may often involve the manipulation of a scene’s lighting. Thus the context for this investigation is based in artistic practice, in perception, and in established instructional strategies.  
QUESTION

We were concerned with the observation and analysis of how and when information about light enters the drawing process. We asked whether differences in how a stimulus is lit affect drawing behaviors and strategies, the hypothesis being that, relative to an overhead diffusely lit scene, a direct-lit scene that presents areas of strong contrast would result in a drawing process that privileges attention to light and shadow. We first conducted a small-scale pilot study to help hone in on appropriate methods. 

The Pilot 

​The two participants in the pilot study were advanced-level drawing students. The stimulus was the bone that was re-used in the main study (see Figures 2a, 2c). The lighting conditions, which were administered in separate sessions, were general, overhead (fluorescent) light and directional incandescent (spot) light. The participants were provided with 6B (very soft) pencils and an eraser. The drawings were done within a 1-week time interval. The participants commented that the stimulus was too complex for a 10-minute drawing, that the conditions created by the directional light were not experienced as significantly different from that of the general light, and that the use of pencil seemed to suggest a need to privilege contour rather than value-driven drawing behavior. As described below, these comments figured prominently in several of the methodological choices we made in the main study.

​For more information on the pilot study, please click here.
PARTICIPANTS, APPARATUS, AND METHOD

​Five participants were tested separately—one professional artist (a drawing instructor for whom drawing from observation is a regular and significant part of her own artistic practice), and four art students who draw regularly but who have different levels of experience. In the case of the students, “experience” was defined in terms of postsecondary drawing courses completed. By including a professional artist, the researchers hoped to witness and measure an experiential differentiation much as described by Solso (2003) in that “while a novice might have to study the specific features . . . in order to tender a drawing of it, the expert would gather pertinent information in a moment’s glance and then allocate further attention to ‘deeper’ cognitive-perceptual matters” (p. 151). 

All participants were right-handed and came to the drawing task with corrective eyewear appropriate to the task of drawing from observation. In ascending order of experience: 
  1. Right-handed female; 17–20 years; contact lenses; no university courses completed
  2. Right-handed female; 21–25 years; contact lenses; one first-level university drawing course completed
  3. Right-handed male; 21–25 years; glasses; two first-level university courses completed
  4. Right-handed female; 21–25 years; contact lenses; multiple university courses completed
  5. Right-handed female; 40+ years; glasses; professional artist
Picture
Figure 1. Drawing setup.

Each participant was seated at an easel, 120 cm from the stimulus. The stimulus was viewed through a black “window” frame with an opening size of 25 cm wide x 20 cm high. The frame was mounted vertically close to midway between the participant and the stimulus (50 cm from the participant and 70 cm from the stimulus; see Figure 1). Use of a window was intended primarily to align, to stabilize, and to standardize participants’ viewing positions. 

For the drawings where direct, incandescent sidelight was used, the window frame also served to hide the light source. Unlike in the pilot study, the drawing surface was left blank. In the pilot, a rectangle proportional to the window through which the participants viewed the stimulus was pre-drawn on each sheet of paper, but pilot participants commented that the presence of a pre-drawn frame was confusing rather than helpful. Also, the drawing tool provided was changed from 6B pencil in the pilot study to willow charcoal in this study. This was because, even though a very soft pencil provides the flexibility to shade, the participants reported that the pencil focused their attention on line rather than on areas of light and shadow. Gum and kneadable erasers were provided in both studies. 

The two lighting conditions used were general, overhead fluorescent lighting (which we refer to as “indirect lighting”) and strong, directional (left) sidelight produced with a Kodak Carousel projector (which we refer to as “direct lighting”). In the debriefing following the pilot study, where a simple spotlight was used for directional light, it was reported by participants that the difference between the types of lighting was not “noticed.” In the main study, the projector replaced the spotlight and produced a much brighter light and, therefore, much stronger shadows than was possible with the spotlight used in the pilot. 

In the pilot, the stimulus was a natural but complex form—a jawbone (referred to as Bone, Figures 2a, 2c)—and, while the researchers considered it appropriate to the task of completing a 10-minute drawing, participants commented negatively on the complexity of the Bone as subject matter for a drawing of limited duration. In the main study, which also used this stimulus, a less complex form was added as a second stimulus (Figures 2b, 2d). Referred to by the researchers as the Braque, it served as both complement (size and shape) and contrast (materials and complexity) to the Bone, and doubled the number of drawing events available for analysis. The stimuli were presented on a white horizontal surface and viewed against a white curved (no edges or corners) background. In Figures 2a and 2b, the stimuli are lit directly from the left; in Figures 2c and 2d, they are lit indirectly from above. 

Prior to commencement of drawing, the stimulus was obscured from view by a white board placed between the black window and the stimulus. Located at the center of the white board was a “target” that aligned to the placement of the stimulus. This provided a focal point by which, following a set of written instructions provided to them (see Appendix A), participants could center themselves to the frame in anticipation of beginning the drawing task. When a participant indicated readiness to begin the task, the white board was removed, video documentation was initiated, and a stopwatch started to time the 10-minute drawing session. 

Over the course of two drawing sessions, each participant was presented with four scenes and completed four drawings: Bone lit by direct light from the left side (Figure 2a); Braque lit by direct light from the left side (Figure 2b); Bone lit by indirect light from above (Figure 2c); Braque lit by indirect light from above (Figure 2d). 

As indicated in the written instructions, participants were notified at three time markers: 5, 8, and 10 minutes. The intention in announcing these markers was to assist the participants in pacing the drawing process, thereby minimizing feelings of being rushed that were noted following the pilot study. 
Picture
​Figure 2. Examples of lighting setup. Lighting conditions are separated by rows (top = direct; bottom = indirect); objects are separated by columns (left = Bone; right = Braque).
Picture
Figure 3. One participant’s drawings. Lighting conditions are separated by rows (top = direct; bottom = indirect); objects are separated by columns (left = Bone; right = Braque).

​On completion of the first drawing, each participant was asked to leave the Drawing Laboratory while a second scene was arranged. On each occasion, the scene-change involved both stimulus and lighting. Returning to the laboratory and making a second drawing completed the first session, and the participant was then invited to return in 7 days for the purpose of making two more drawings. Nothing was revealed about the nature of the second session, and participants were asked not to discuss the first session. When the participants returned to the laboratory after 7 days, the process was repeated, the difference being that the lighting for each stimulus was changed. Minor changes were also made to the orientation of the objects for the purpose of adding a second, subtle level of visual novelty. We surmised that participants would be less inclined to resort to memory-drawing if they thought that the relationship of the object to their viewing position may have changed. 

Upon finishing the fourth drawing, each participant was asked to complete a short questionnaire (see Appendix B).
ANALYZING THE DATA—DRAWINGS, DIGITAL RECORDS, AND QUESTIONNAIRES

Drawings 

For purposes of analysis, we looked first at the finished drawings (for illustrative purposes, four drawings from one participant are shown in Figure 3). These were grouped and reviewed side by side. This was done first by participant, second by subject, and finally by lighting. 

Predictably, each participant’s set of finished drawings exhibited similarities in mark, composition, and style. That is, each set of drawings possessed common qualities. Taken alone, similarity in the appearance of end products may suggest that a hypothesis wherein difference in stimulus-lighting results in different approaches to the drawing process is invalid. 
Picture
Figure 4. Drawing generated by one of our participants with the impact, on this drawing, of the light depicting behaviors coded from the video recordings noted.

However, the digital record provided the opportunity to review the evolution of each drawing and to witness when attention to light was significant within the process itself. 

Digital Records 

Analysis of the digital record entailed each 10-minute drawing event being viewed in real-time. This was done in two stages. First, the development of each drawing was viewed individually. Next, drawings made from the same stimulus were compared side by side on a single screen. 

Furthermore, each video was reviewed with the 10-minute drawing period divided into sixty 10-second epochs: that is, each 10 seconds was marked off during viewing. Only when the participant’s hand obscured the image at a 10-second time-marker was the video moved forward or back to reveal the drawing surface. The particular behavior that dominated during each epoch was recorded. Three distinct light-depicting behaviors were identified: (a) establishing tonal variation by shading, (b) varying line quality to suggest differences in light and dark, and (c) erasing to lighten a particular area of the drawing. Although these behaviors were derived from the dynamic behaviors observable in the video recordings, examples from a static image are presented in Figure 4. 

Differences in light and dark entailed, for light, both the effects of direct light and of reflected light. For dark (shadow), three kinds of shadows, as described by Jacobson and Werner (2004), were considered: grounding shadows, which serve to indicate that the object is located on a surface; attached shadows, the shadows on an object that convey form, through an activity like shading; and cast shadows, shadows from one object that are cast upon other objects in a static image. 

Non-light-related mark making behaviors were classified as “other.” Examples of other behaviors are enveloping (completion of an outline), depicting background information, and erasing to correct mistakes. 

The behaviors recorded for each 10-second epoch for each participant and each object under the two lighting conditions constituted the data that were subjected to statistical analysis. Although we had a relatively small sample size, with five participants and four drawing conditions, we were still able to conduct inferential statistics to test our hypothesis that one or more of the three distinct light-depicting behaviors were more frequent with direct than with indirect lighting (see “Results,” below). It is also worth noting that small sample sizes characterize other studies of drawing behavior. For example, Miall and Tchalenko (2001) used only one artist, while Solso (2001) compared one novice with one professional. To minimize the probability of a type II error, one-tailed t-tests were conducted on each of the three light-depicting behaviors. Two-tailed t-tests were conducted for all other comparisons. 

Questionnaires 

​The purpose of the questionnaire was twofold: first, to contextualize our understanding of individual processes by asking participants to reflect upon the experience of drawing, and, in particular, to describe any drawing strategies that may be influenced by changes in lighting; and second, to ask about the design of the task and, for future reference, to seek advice as to how similar tasks could be better designed and organized.  ​
RESULTS
​
The results of our analysis of the light depicting behaviors described above are illustrated in Figure 5. With regard to our hypothesis, the overall rate of tonal drawing was significantly greater (t4 = 2.74, p < 0.05, one-tailed) for direct lighting (46%) than for indirect lighting (38%). There were no consistent effects of lighting on the other two light-depicting behaviors. 

There were interesting results that were dependent on the nature of the stimuli. Compared to the Braque object, there was more light-depicting behavior for the Bone: the overall rate of erasing to create light(8% vs. 4%) was significantly greater (t4 = 2.94, p < 0.05, two-tailed), and the overall rate of line related tonal drawing (7% vs. 4%) was marginally greater (t4 = 2.45, p < 0.10, two-tailed) for the Bone.
Picture
​Figure 5. Aggregate activity over the 10-minute drawing period by category. Lighting conditions are separated by rows (top = direct; bottom = indirect); objects are separated by columns (left = Bone; right = Braque).
DISCUSSION

Although statistical analysis indicates that the hypothesis (that relative to a diffusely overhead-lit scene, a directly side-lit scene that presents areas of strong contrast results in drawing processes that privilege attention to light and shadow) was essentially correct, we were surprised at the consistency in drawing behaviors with which each participant approached each different lighting situation. There was an expectation that the more strongly lit, tonally dramatic scene would result in a different initial approach to the drawing, one that foregrounded attending to tonal variation. This was not the case. Instead, each participant began each of his or her drawings using a very consistent approach. There were, of course, differences between the drawing techniques of the participants, but it was also clear that the existing drawing strategies brought to the study by each participant had more impact on the drawings than did variations in lighting. Interestingly, it may be that familiar strategies actually impeded the ability to attend to tonal variation. As suggested by Hochberg (1979), “[s]chemas . . . loosely equivalent to images” (p. 18) are used whereby observers continuously fit the schemas that they bring to a situation to what is given by each momentary glance. It is conceivable that a similar dynamic occurred with our participants, whether their schema is a general image of what the drawing should look like when complete, or a set of learned strategies that are brought to bear on each new observational drawing task. In a study designed to make use of significant differences in lighting, the influence of schemas may have resulted in a tendency to overlook lighting differences in visual stimuli. 

The decision to include a second stimulus was intended to address the concern, raised by the participants in the pilot study, that the Bone was considered to be “too complex” to draw completely in 10 minutes. The Braque had a smooth finish and six distinct elements that either receded into or protruded from the general form. It is interesting to note the difference in drawing strategies evoked by the Braque. One explanation may have to do with the fact that the discrete elements in the Braque suggested themselves to be more easily described in graphic terms, thus appearing to be more manageable for the participants to draw. Rather than generalize, as they did with the Bone, participants attempted to portray each element of the Braque in its proper place and in proportion within the drawing. This approach requires that the early stages in the drawing process be more precise, including erasing to correct rather than “illuminate.” As a result, less drawing activity was directed toward an understanding of the effects of lighting. This approach would seem consistent with the notion that:
our perception tends under natural circumstances to be ‘object-directed’ rather than ‘stimulus-directed.’ That is, we customarily seek to achieve an accurate perception not of the isolated stimulus attribute, but of the whole object. The functional value of this is obvious, inasmuch as it is with whole objects in our environment that we must cope. In order to see objects, we must necessarily take account of patterns of stimuli, not isolated stimulus attributes, and in so doing we can achieve perceptual constancy of objects. (Krech et al., 1969, p. 215) 
Finally, responses to the question “Did you approach the drawing differently because of the difference in lighting?” suggest that, even though the behavioral and physical data (the final drawings) appeared to only mildly support the hypothesis that lighting makes a difference, the participants reported a more significant level of awareness and of application:
  1. “I did. The (side) lighting created more depth and shadows so it was easier to approach.”
  2. “I think so. Strong shadow in skull a starting point.”
  3. “Yes—preference toward strong lighting: exaggerated side plane to create form/drama.”
  4. “Slightly, I think having stronger shadows allows you to see more individual shapes within the objects and allowing you to use those shapes to find tonal relationships that help define the object, and I think it also allowed me to include tone earlier in the process.”
  5. “Yes—I was more concerned with the total page because of the strong cast shadows when under the directional spotlight.”
​
​Drawing instructors who invest time and energy in carefully lighting a scene can be encouraged by the fact that, even though changes in lighting may sometimes result in only subtle differences in drawing strategies, having one’s attention drawn to the effects of light on a scene does appear to translate into behaviors that are both practical and useful. Nicolaïdes (1969) is correct when he observes that “[l]ight and shadow can be a very real help to you in seeing and understanding the form. . . . Shadows can be a helpful element only when you make the integrity of the form your first consideration” (p. 140).
IMPLICATIONS FOR FURTHER STUDY

  1. In order to strengthen the statistical findings, a larger sample of participants could be used.
  2. A greater variety of objects could be used as drawing subjects. Such objects could have surfaces that expand the type of visual pattern that drawers may encounter. These objects would need to be constructed rather than found, so that they remain unrecognizable to participants, thereby reducing the opportunity to apply familiar and pre-existing schemas. The different surfaces could have various levels of complexity in order to explore the phenomenon that the researchers identified between the need to generalize, as in the case of the Bone, and the desire to more accurately represent, as demonstrated in drawings of the Braque.
  3. Materials selection could be left to the participants. Even though a change in drawing materials was made from pilot to main study, one participant commented that the charcoal was a limiting rather than accommodating medium. This was similar to our view of the drawing tools used in the pilot where pencils seemed to “suggest” line rather than area to the participants. Even though it would introduce a new variable, a more open approach to materials choice may be less inhibiting to the participants.
  4. Further research regarding the application of drawing schemas could test the consciousness of use, the frequency of application, and the degree to which such behavior may be modified through teaching. 
APPENDIX A: INSTRUCTIONS GIVEN TO PARTICIPANT

PRIOR TO STARTING THE TASK OF COMPLETING A TEN-MINUTE DRAWING, PLEASE ENSURE THAT: 

The target board is centered (left to right) in the black window frame; and 

You are at a comfortable distance from the drawing surface. 

When the target board is removed, you will see the scene to be drawn. 

FYI: 

Charcoal and erasers are provided. 

You will be notified at the 5- and 8-minute mark. 

​Your drawing will be retained for research purposes. 
APPENDIX B: QUESTIONNAIRE ADMINISTERED TO PARTICIPANTS

THE FOLLOWING ARE QUESTIONS THAT ARE SPECIFICALLY ABOUT YOUR RECENT EXPERIENCE DRAWING IN THE LAB. 
  1. Can you describe the difference between the experiences of making each drawing?
  2. Did you approach the drawing differently because of the difference in lighting?
  3. Can you comment on the subject matter as it may have influenced how you worked?
  4. To what extent do you think that the drawing tools influenced process?
  5. As much as we hoped that the setup presented to you would feel like a familiar and comfortable drawing situation, was there anything about the setup that changed how you would typically draw from observation?
  6. Did memory affect how you approached the second drawing session?
REFERENCES

​Enstice, W., & Peters, M. (2003). Drawing: Space, form, and expression (3rd ed.). Upper Saddle River, NJ: Pearson Prentice Hall.

Goldstein, N. (1999). The art of responsive drawing (5th ed.). Upper Saddle River, NJ: Prentice Hall.

Hess, E. H. (1961). Shadows and depth perception. Scientific American, 204(3), 138–151.

Hochberg, J. E. (1979). Some of the things that paintings are. In C. F. Nodine & D. F. Fisher (Eds.), Perception and pictorial representation (pp. 17–41). New York, NY: Praeger.

Jacobson, J., & Werner, S. (2004). Why cast shadows are expendable: Insensitivity of human observers and the inherent ambiguity of cast shadows in pictorial art. Perception, 33(11), 1369–1383.

Krech, D., Crutchfield, R. S., & Livson, N. (1969). Elements of psychology (2nd ed.). New York, NY: Knopf.

Miall, R. C., & Tchalenko, J. (2001). A painter’s eye movements: A study of eye and hand movement during portrait drawing. Leonardo, 34(1), 35–40.

Myers, J. F. (1989). The language of visual art: Perception as a basis for design. Fort Worth, TX: Holt, Rinehart and Winston.

Nicolaïdes, K. (1969). The natural way to draw: A working plan for art study. Boston, MA: Houghton Mifflin.

Shlain, L. (2007). Art & physics: Parallel visions in space, time, and light. New York, NY: Harper Perennial.

Solso, R. L. (1994). Cognition and the visual arts. Cambridge, MA: MIT Press.

Solso, R. L. (2001). Brain activities in a skilled versus a novice artist: An fMRI study. Leonardo, 34(1), 31–34.

​Solso, R. L. (2003). The psychology of art and the evolution of the conscious brain. Cambridge, MA: MIT Press.