Handwriting-movement control: research into different levels of the motor system
Hans-Leo H. M. Teulings · Radboud Repository (Radboud University) · 1988
Chapter 1 2. Memory RetrievalFor the information stored in motor memory to be actually used in a motor program, the appropriate movement information has to be retrieved.However, it seems unlikely that each writing pattern is completely represented in motor memory.It is more likely that only separate strokes or sequences of a few strokes are represented as ready-to-retrieve units of movement.In Chapter 3 a paradigm will be presented, which allows us to discriminate between the hypothesis that a single stroke and the hypothesis that a multistroke allograph forms a unit.In the latter chapter it is argued that congruent strokes, i.e., different strokes having the same turning direction, may be represented in motor memory by the same unit.Now, three types of letter pairs can be constructed: pairs of identical allographs, pairs of different allographs but constructed from congruent strokes, and pairs of different allographs constructed from noncongruent strokes.It is then possible to distinguish between these two hypotheses.As the memory-retrieval process takes place whenever a new unit is required, it will mainly affect movement latency (i.e., choice reaction time).It appears that pairs of identical allographs yield short choice-reaction times whereas pairs of different allographs, no matter whether their strokes are congruent or noncongruent, yield long choice-reaction times.These data support the notion of complete allographs being represented as units in a long-term motor memory.Why is cursive handwriting so fluent, whereas it appears to consist of a sequence of discrete units?It is probably not true that movements are executed immediately after retrieval from motor memory.Instead, the entire movement sequence has to be prepared or organized prior to, or during, movement execution.Various authors suggested mechanisms of preparation.For instance, Sternberg et al. (1978) and Ellis (1982) suggest that the movement sequence has to be loaded into a buffer first.Rosenbaum et al. (1984) suggest that a hierarchical tree has to be set up.Whatever the preparation mechanism is, it takes time.Hence, if a subject can prepare a sequence prior to a 'go signal', movement latency (i.e., simple reaction time) will be much smaller.However, more interesting is that both mechanisms imply specific processes during movement execution, namely retrieving movement information from a buffer, or traversing a hierarchical tree, respectively.Therefore, one may expect that the structure of the units in the writing pattern may affect movement duration.A known but unexplained effect is the observation that sequences of identical units take more time to execute than sequences of different ones.In our investigation, pairs of HSR.