Topics

    Drawing and the blind:  John M. Kennedy



1.   Could outline function in touch?

2.   Stages:  Vision uses brightness differences - but lines
     can be black or white!

3.   The explanation of outline will NOT be "edges and
     lines have contours in common".  (Rather, it depends
     on "axes".

4.   The "house and hills" drawing shows relief changes.

5.   These relief changes operate without training, in
     infancy (Hochberg and Brooks; DeLoache).

6.   They operate across cultures (Kennedy & Silver,
     Kennedy & Ross).

7.   The changes are depicted by LINES with
     JUNCTIONS.

8.   "Chiaroscuro" or "shading" provides impressions of
     shape i.e. "shape from shadow".  It operates via
     gradual or abrupt borders or "contours" on surfaces.

9.   Lines generally have two contours on surfaces,
     providing optic contrasts.



10.  Contrasts can come from corners or stains or
     shadows.  Therefore, they are ambiguous about their
     origins.

11.  The observer uses information to "disambiguate"
     contrasts.  The shapes of contours include several
     types - familiar ones, perspective, topological, fractal
     etc.

     ASIDE - N.B. geometries include CONGRUENCE,
     SIMILARITY, SHEARING, PERSPECTIVE,
     TOPOLOGY.  Fractal shapes are SCALE FREE.

12.  Hypothesis:  Can an outline substitute for any
     contrast?

13.  Figures with abrupt shadow borders are often
     identifiable.

14.  They give 3D shape-from-shadow.

15.  Most of their borders can be shadows on continuous
     surfaces, not just shadows at corners.  Yet outlines
     based on these figures are very hard to recognize.

16.  Only minor loss of recognition, if any, is suffered by
     b/w figs if the b/w patches are NOT due to shadow &
     illumination see STREET'S FIGURES - Fig. 2.6






17.  Outline fails with chiaroscuro in two ways.

     -    firstly, we cannot see the object with shadowed &
          illuminated regions (we can't do shape-from-
          shadow analysis).

     -    secondly, we can't see regions we take to be
          "shadow" shapes as looking darker than regions
          we take to be "illuminated" regions.

18.  Some related studies

     Kennedy & Silver - Caveat, no lines for chia.

     Kennedy & Ross - Songe fail to recognize line for chia
     - colour change.

     Fussell & Haaland - Nepalese  fail to use outlined
     regions that worked as cross-hatched regions.

19.  Curiously, ends of lines do give brightness effects.

20.  In Street figures, outlines do not work like patches of
     solid colour - because ------> "Scission" only occurs
     with solids.

21.  Outlines tracing shadow boundaries don't become
     dark in appearance on one side when we realize what
     they are copying.

22.  However, outlines tracing corners or edges do provide
     changes in the appearance of surface relief when we
     realize what they are copying.

23.  Outline can function with very schematic pictures. 
     Evidence:  Kennedy and Ross.

24.  How is visual outline like tactile outline?

     A.   Outline and touch both use relief.

     B.   Outline in touch could use ridges, in place of
          visual contours.

     C.   Visual outline uses axes - so too could touch.
          (Axis are NOT brightness phenomena).

     D.   Outline can be schematic - useful in touch.

     E.   Outline works without training in vision.  Could
          be the same in touch?

25.  Machinery

     Do lines stand for "high spatial frequencies"?  Hayes
     said "the low spatial frequencies are the reason for the
     loss" in recognition of negatives i.e. the locations of
     large expanses of black and white.  But --- this just
     redescribes the problem!


26.  Is a shadow always" within a range of frequencies"? 
     That is, is it always a certain MULTIPLE of the
     object casting the shadow?

27   How do we find the axis?  Midway between contours? 
     This is often true, but not always.  (Rubin noted 2
     close contours forming a line do not have congruent
     contours!)

28.  We could "extract" the axis, and "omit" information
     about where the line is dark or bright.  (This could
     help explain why we use both dark lines on white
     surfaces and also white lines on dark surfaces).

29.  We may have axis-extraction systems in some visual
     channels but not others.  

     N.B. Axis extraction may cooperate with "subjective
     contours" and "scissions".  These do not exist in the
     colour channel.  And scission is impossible in outlined
     areas.  Also, the motion channel is rather "crude" or
     "insensitive to high spatial frequencies".  The ITS
     channel is sensitive to detail.  It may use some of
     visual area V2 to create subjective contours.  It may
     have several functions, including outline, working in a
     cooperative fashion.  

30.  Evidence:  Rose, Gottfried and Bridger (1983). 
     Infants touch objects.  Then they can recognize them
     visually, as outline drawings.  Let's follow this up -
     with the blind!






          
 

This page was designed and is maintained by William Barek, CITD - Bladen Library.
E-mail: barek@macpost.scar.utoronto.ca