Abstract <p>Previous conversation analytic studies of instruction have observed that instructors use deictic expressions, accompanied by gestures and other body movements, to demonstrate how to complete a task. In most examined cases, these deictic expressions are proximal and refer to the object of the speakers’ (i.e., the instructors’) pointing gestures or their body movements themselves. By contrast, in this study, we focus on uses of the Japanese medial (<i>so</i>-type) spatial deictic pronoun (<i>soko</i> or “that place”) in instructive actions to explore aspects of the organization of instruction and instructed action. We observe that this medial deictic expression refers to the site of the recipient’s (i.e., the novice’s or student’s) specific operation or their current posture. We use conversation analysis to analyze videotaped interactions in instructional settings. We demonstrate that the instructors incorporate the novice’s or student’s action of following a just-given instruction, or its result, into an incipient follow-up or next instruction by using the medial deictic expression. We also discuss some theoretical implications of the demonstration by highlighting the distributed nature of the medial deictic reference and the “co-operative” nature of instruction.</p>