We shall make no attempt to answer each of Schrier and Thompson's (1984) many criticisms of our report (Menzel & Juno, 1982). Instead, we shall first list some of the points on which we actually concur with these authors (usually despite their apparent opinions to the contrary), and then we shall expand upon a few details mentioned in our report, which they overlooked. If any loose ends still remain, we request readers to consult, or reread, our original report and Kamil and Yoerg's (1982) overview of the relations between learning and foraging behavior.The points of agreement are as follows: (1) Our report is not directly concerned with interproblem improvement in performance during the course of our observations. (We could detect no such improvement. Insofar as direct comparisons of data are warranted, our animals seemed to start where marmosets might be expected to asymptote. And if we must demonstrate interproblem improvement before talking about "learning sets," is not the doctrine that learning sets are learned unfalsifiable?) (2) Our report does not contain conclusive evidencethat the difference between our data and data from "real" learning-set studies is attributable to testing the animals in their home cage as a group instead of in isolation. (3) The ability for one-trial learning in our subjects does not prove that the genesis of this ability involved no trace of slow, gradual learning at any point in their lifespan. (4) It is hard to imagine that test-naive primates of any and all species would perform as our animals did. (5) Our methods are so different from those of Harlow that precise comparisons of the numerical data would be of dubious validity. (6) Students of learning set should not feel compelled to adopt the methods and concepts of foraging theorists if they do not want to. (7) The data we presented do not, as such, prove that animals were cognitive geniuses. (8) Slow, gradual learning is not necessarily maladaptive or unknown outside of the laboratory. (9) Our animals did not learn all that could conceivably be learned about an object in a single trial. were a single individual. And, coincidentally, we also presented data that we construed as strong evidence of learning on the part of individual group members. Schrier and Thompson do not mention these data.) (13) Laboratory research is a good thing and can teach us a lot that we could not learn otherwise. (14) Readers who did not read the references that we cited might possibly infer, incorrectly, that learningset theorists today still hold that performance on learning-set tasks is a good index of so-called general intelligence. (15) We did not provide a comprehensive review of the history or theory of learning set.The principal purpose of our report was to describe, for the benefit of scientists in general, rather than only specialists in learning, how group-living Saguinus fusctcollis behave with respect to one another and with respect to novel objects. It extended a previous study on this same general topic (Menzel & Menzel, 1979) and al...