Showing papers on "Naturalness published in 1985"

About natural logic

Abstract: The use of first-order logic in data bases raises problems about representing relations that are not defined everywhere. The solution to this problem is to use a 3-valued logic with the truth value “undefined”. This paper is concerned with the naturalness of such a logic, its power and its connection with classical logic.

Measurement and Modification of Speech Naturalness during Stuttering Therapy

Abstract: This paper describes two studies that illustrate the utility of listener ratings of speech naturalness for measuring and modifying speech naturalness during a stuttering therapy program. The progra...

Modification of Listener-Judged Naturalness in the Speech of Stutterers

Abstract: This study investigated the effect of regular feedback of listener-judged speech naturalness ratings on the speech of stutterers. Six adult stutterers each participated in a time-series ABA experiment. During the treatment phase the stutterer was instructed to improve a clinician's rating, on a 9-point scale, of the naturalness of each 30-s interval of the stutterer's spontaneous speech. The results indicate that the naturalness ratings and stuttering for 5 of the subjects made favorable changes during the treatment phase. Analyses of the findings show that only some of the naturalness judgments were influenced by stuttering frequency and speech rate. A perceptual analysis of the speech of 2 subjects suggested that the speech naturalness ratings were also probably influenced by other less obvious variables.

On Logo Squares, Triangles and Houses.

Abstract: and is fixed up by an initial RT 90 command. Papert then discusses the significance of the above episode by pointing out how the use of procedures as "building blocks" is embedded in the idea of hierarchical organization, and how the activity of debugging in LOGO is often incorporated into the process of understanding. I will be describing below four hour-long sessions of children working with the definition and use of a procedure for triangle. Now, I don't know how old Pamela is, since Papert doesn't tell us (indeed, throughout Mindstorms, it is not always obvious when Papert is talking about S year olds and when he is talking about 15 year olds). But for the 9 year olds that I have been observing, the sequence (program + bugs) -> debugging new understanding, portrayed above, does not even begin to capture the level of conceptual difficulties experienced by these children, nor, for that matter, does it capture the richness of the experience. The four children 111 be describing would be in their 26th to 30th session of turtle-geometric activities. Their "LOGO environment" is somewhat privileged both in having a complete access to computers (in the computer lab of the math department) and in being provided with a lot of help, ideas and challenges. On the other hand, since our research is focused on the acquisition of mathematical and programming concepts, we have often structured activities to suit our research agenda rather than the children's immediate interest (I recognize that, in the extreme, such an approach is antithetical to the whole LOGO philosophy. But the viewpoint that adults have nothing to offer to children is equally absurd.) For more details on the first 12 sessions, see Hillel [1984]. I would like to first examine the two important ideas mentioned by Papert, namely, the idea of "hierarchical organization" and that of "debugging". Clearly, defining and using LOGO procedures as "building blocks" is central to the programming activity. This involves both the synthetic activity of using existing procedures to construct more elaborate figures (as in the example recounted by Papert) and the analytic activity of making a "proceduralanalysis" of, say, a complex figure in terms of identifiable "building blocks" (subprocedures). Both kinds of activity ought to foster the idea of "hierarchical organization". Yet the children we observe and those observed by others (see for example, Leron [1984]), do not spontaneously adopt a procedural viewpoint when writing a program. Despite the fact that our children have been provided with extensive experience of using procedures and the fact that we have discussed the relative merits of using procedures, they, surprisingly often, opt for writing simple programs (i.e., using only LOGO primitives). I think that a plausible explanation for this is the overwhelming identification by the children of programming as "drawing with the turtle". On the one hand, part of the "naturalness" of LOGO is due to the strong resonance of "instructing the turtle" with the children's drawing schema. On the other hand, linking programming with drawing stands in the way of making a "procedural-analysis" of a particular task. Writing a program to produce a geometric figure is viewed as tracing certain linear or curvilinear segments rather than putting together some, possibly not yet defined, building blocks. I do not dispute Papert's claim about the importance of debugging to the learning process, but I do question whether a lot of the children's attempts to fix their procedures should be called debugging rather than "adjustment