A return to basic inductive research methods is recommended. Specifically, the scientific method of "strong inference" ought to be applied in attempts to prove hypotheses concerning the relationship that may exist between attitude variables and the learning of a second or foreign language. These hypotheses, drawn from the literature and discussed here, are: (1) the strength of the relationship between attitudes and achievement increases with increments of time; (2) the relationship becomes weaker under the same conditions; (3) the relationship tends to remain unchanged; (4) the relationship between attitude variables and variance in second language learning is quite strong; variance which could be accounted for by attitude variables might be above 25%; (5) the variance might be much less, below 10%; (6) the relationship may be substantially stronger in contexts where many opportunities to communicate with the target language group are available; (7) the more a learner is self-confident, extroverted, and friendly, the more rapid will be his progress and the higher his ultimate level of attainment of proficiency; (8) the direction of causation is from attitudes to learning and achievement; (9) high levels of attainment or rapid rates of learning may cause positive attitudes; (10) the direction of causation is both from attitudes to attainment and the reverse; (11) the relationship of attitudes and achievement will be the same in foreign and second language learning contexts; (12) the contrast will reveal a stronger relationship in foreign language contexts; and (13) an integrative orientation will produce more efficient learning and a higher level of attainment than an instrumental orientation. (Author/CFM)