One day an ant strayed across a piece of paper and saw a pen writing in fine, black strokes.

"How wonderful this is!" exclaimed the ant. "This remarkable thing, with a life of its own, makes squiggles on this beautiful surface, to such an extent and with such energy that it is equal to the efforts of all the ants in the world. And the squiggles which it makes -- these resemble ants; not one, but millions, all run together.

He repeated his ideas to another ant who was equally interested. But another ant said, "I have now also observed this strange object. But I have determined that it is not the master of this work. You failed to notice that this pen is attached to certain other objects which surround it and drive it on its way. These should be considered as the moving factor and given the credit." Thus were fingers discovered by the ants

But another ant, after a long time, climbed over the fingers and realized that they made a hand which he thoroughly explored by scrambling all over it. He returned to his fellows, "Ants! I have news of importance for you. Those smaller objects are part of a large one. It is this which gives motion to them."

But then it was discovered that the hand was attached to an arm, and the arm to a body, and there wre feet which did no writing. The investigations continue. Of the mechanics of writing the ants have a fair idea. Of the meaning and intention of writing and how it is ultimately controlled they will not find out by their customary method of investigation.

• symbolic drawing is uniquely human activity
• sign of civilization
• in Papua New Guidea, small tribe of Songe
• no symbolic representational drawings
• although they could recognize drawings when shown
• The planning of strokes
• General rules
• several general consistencies or rules have been observed
• start at leftmost point
• start at the top
• start with a vertical line
• for figures with an apex, start at the top and descend the left oblique
• Draw horizontal lines from left to right
• draw vertical lines from top to bottom
• keep the pencil on the paper at all times
• Investigate children
• Goodnow & Levine (1973)
• analyzed drawing paths that children followed when copying 2-d shapes
• with development the children began to follow the observed rules
• Learning to draw may be path on road to learning to write and speak
• evidence from agrammatic aphasia
• impairment in forming sentences
• unlike hearing or deaf but otherwise normal children, these did not copy figures in systematic fashion
• possible problem in generating appropriate sequences of behavior
• Benefits of general rules
• simplify motor planning
• promote flexible performance while reducing memory storage
• satisfy biomechanical contraints, e.g., avoid extreme joint angles
• Top-down influences
• Stroke patterns influenced by semantic or perceptual interpretation
• The isogony principle
• Equal angles are described in equal times
• Complex curves segmented into components
• Timing is primary control variable
• Slowing on tight curves
• Two-thirds power law
• Angular velocity(t) = k curvature(t)^2/3
• Scaling of k with movement amplitude
• Hypothesized mechanism: coupled oscillators
• Drawing smoothly
• Examples of departures from 2/3 law
• Question biological plausibility of sinusoidal oscillators
• Hypothesized mechanism: minimize jerk
• can give rise to 2/3 power relation

• Error analysis
• Types
• Lexical
• Letter substitution
• Letter repetition
• Letter substitution and case change
• Extra stroke
• Hypothesis that writing planned in stages
• word grapheme allograph graph
• Word stage is selection of what word to write. Error at word stage leads to lexical errors
• Grapheme stage is selection of letters. Error at grapheme stage leads to substitution and repitition errors
• Allograph stage is selection of grapheme's categorical variations, e.g., upper or lower case. Error at allograph stage lead to case changes.
• Graph stage is selection of strokes. Error at graph stage lead to stroke errors.
• Agraphia
• Impaired writing but preserved motor control, perception, spelling
• usually occurs with other impairments due to extent of brain damage
• Patient with normal retrieval of word, grapheme and allograph but impaired stroke production
• Another patient identified with selection deficit in grapheme selection
• Reaction-time evidence for grapheme selection (stage 2)
• Teulings et al (1983)
• Method
• Display handwritten letter on screen
• Present another letter to left or right
• Ss copy letter pair as soon as possible
• Result
• Time to start writing was shorter when the two letters matched than when they did not match
• Interpretation
• Ss were faster when single grapheme was selected. Additional time was needed if second grapheme was required.
• Reaction-time evidence for allograph selection (stage 3)
• Stelmach & Teulings (1983)
• Method
• Ss wrote one of two letter strings but one string was more likely than the other. By making one string more likely, Ss presumably were more likely to prepare writing it so that if the other string was called for, some time was needed to switch the readied program.
• Two choice
• different letter shapes - hye and ynl
• different letter sizes - hye and hye
• Result
• in different letter shape condition the time to start writing the improbably string was 128 ms longer than to start writing the probable string
• in different letter size condition the time to start writing the improbably string was 63 ms longer than to start writing the probable string
• Interpretation
• RT cost was less in letter size than in letter shape condition because letter size specified after letter shape
• Writing size, relative timing and absolute timing
• How is writing size governed?
• Possibly a single parameter like timing controls the size of the strokes - rate scaling. Evidence consistent with rate scaling
• Observed that larger letters take longer to write [this may seem obvious, but it doesn't necessarily have to be that way]
• Nitrous oxide inhalation produces larger writing strokes, presumably because brain processes slowed down
• I'm not going to mention the hypnosis study!
• Observation that spontaneous changes in writing rate result in changes in writing size
• Viviani & Terzuolo (1980) measured writing as subjects wrote the same letter several times - "a"
• Plot tangential velocity and arrange from shortest duration to longest duration, i.e., smallest to largest letter
• The profile has certain reversal points
• A straight line can be fit to the times of the reversal points across all movement durations
• This means the ratios of successive stroke durations remained constant
• Thus, changes in letter size result from absolute change in stroke duration but not relative times within strokes remain constant
• Letters written by different hands appear to maintain the same relative stroke duration ratios. Differences in handwriting appearance partly attributable to differences in absolute intervals
• Machines using these principles can simulate handwriting!
• Context effects
• Shape and timing of a letter depends on what letter precedes and follows it
• ee or ll written faster than el or le
• presumably motor program has to be changed to switch letters
• this finding is consistent with the notions of hierarchical motor programming discussed earlier in class
• Writing and handedness
• Left hemisphere dominance for language, and left hemisphere control of right hand
• Left handed writing may have right hemisphere dominance for language but not necessarily
• Inverted versus noninverted left handed writing. 50% of each.
• Jere Levy proposed that inverted left-handed corresponds to language function dominant in left hemisphere, and noninverted left handed writing corresponds to language function dominant in right hemisphere
• Investigate verbal processing of visual stimuli presented to each hemisphere.
• Found that inverted left handers reported verbal stimuli presented to left hemisphere whereas noninverted left handers reported verbal stimuli presented to the right hemisphere
• Moscovitch presented evidence elaborating on this finding
Hookers & nonhookers did manual RT response with either hand to visual, auditory & tactile stimuli. Results not entirely consistent with Levy hypothesis.