Our study falls into two distinct but related parts.It provides a short critique of current CAI(Computer Assisted Instruction)or CALL(Computer Assisted Language Learning),identifying in particular the pedagogic limitations of many programs and their heavy production costs in relation to their range of applicability.One drawback we identify is the limitations imposed by the restricted processing power of the most microcomputers and the inappropriateness of BASIC for anything other than relatively simple language tasks.As a contrast to its criticism,however,we offer in the second part of our next paper in innovation for CALL:one which does not replicate what many teachers have been doing in their noncomputerized classrooms,but which makes accessible to learners analyzed corpora of language data on which they can draw.Using a powerful database program(Hyper Card),we provice an extensive illustration of work in conventional pragmatics,specifically the selection of sociolinguistically appropriate realizations of chosen speech acts.What this example shows is the potential for such linguistic database in classroom language learning.Not only can learners access very large and varied input data,they can also,in principle,draw on the results of linguistic analysis as an aid to their own problem-solving.Used in this way,CAI or CALL has a liberating effect on the learner,and suggests considerable changes in the relationship between teachers and learner.Indeed,as a result of that,we will see an area of cooperation between CAI and current theories of language learning and classroom pedagogy.|市販のCAIソフトは,教育的な配慮においう不十分なものが多く,その応用範囲の狭さ,また,学生に興味をあまりそそらないということ,に比して高価なものが少なくない。また、パーソナル・コンピュータのもつ処理能力の限界からくる問題,比較的単純なタスクに限定されるプログラム言語BASICの問題等がある。CAI教育の運用面に関しては,受講者1名に対して1台のパーソナル・コンピュータを割当なければ効果的な教育は期得できない。今後の研究課題として,ワン・ウエイ・コミュニケーション的なプログラム追従型の授業形態ではなく,受講生間並びに教師と受講生の間のコミュニケーションおよびコンピュータ人間の間で行なわれるインターアクティブな作業を触発するものを目指したい。