Did Scholar, the TEL environment of the 70s, involve adaptivity or personnalisation?

A recent discussion of the TEL Dictionary initiative LinkedIn group raised the question of the existence of a personnalisation or adaptivity objective in the design of SCHOLAR, one of the seminal TEL environments (Carbonell 1970 ). Indeed it is true that the whole objective was to adapt to the learner in a more flexible way than ever before. This effort was based on two principles.

1. the use of semantic networks ensures that the machine and the learner have similar knowledge structures (ibid. p.197), facilitating a kind of shared understanding.
2. "mixed-initiative dialogue" would allows a better adaptation of the interaction.

Hence we may think of a search for personalisation and adaptivity, although it might not have been the case. Or at least, not the case in these terms. This is well illustrated by the last words of an other report: "what we have tried to show in this paper is the fuzzy, ill-defined, uncertain nature of much of human knowledge and thinking. We want SCHOLAR to be just as fuzzy-thinking as we are." (Carbonell and Collins 1974). So, there is no evidence that the objective was personalisation. Actually after a close analysis of the types of learner's error (ibid. p.198), Carbonell develops a n argument in support of the claim that teachers are less interested in diagnostic than in allowing students to overcome errors. Interestingly enough, noticing that SCHOLAR at that point had limited diagnostic capacity, he writes: "The system could also ask for help when complicated diagnoses appear needed." (ibid. p.200). Unfortunately he died too early (1973) to develop SCHOLAR further, but I would suggest that he was following a different route than personalisation and adaptivity. An objective more related to the modelling of conversation with a knowledgeable other who identifies your errors and drive the conversation to correct them not necessarily diagnosing their origins or making sense of what he or she thought. That would be in line with the more general objective of Carbonell who viewed "the Scholar system as an environment to study natural semantics" (Carbonell and Collina 1974).

Illustration taken from Carbonell and Collins, 1970, "Mixed-initiative systems for training and decision-aid applications" (see the document there). 

Key reference: Carbonell, J. R. (1970). AI in CAI: An Artificial Intelligence Approach to Computer-Aided Instruction. IEEE Transactions on ManMachine Systems, 11, 190-202. IEEE.