Research on Spanish/English code switching is reviewed and the definitions and categories set up by the investigators are examined. Their methods of locating, limiting, and classifying true code switches, and the terms used and results obtained, are compared. It is found that in these studies, conversational (intra-discourse) code switching is separated from situational (stylistic) code switching because the latter is motivated by obvious social factors while the former may have various motivations. Also, intrasentential switching is found to be more interesting linguistically than intersentential switching because of its complexity. In such research, it has usually been found necessary to exclude terms already borrowed by the community. Proficiency, attitude, and education are all seen to affect the type of code switching engaged in, but some common rules are followed. It is concluded that the different results obtained by investigators may be due to different analysis methods and to different ways of classifying switches. A comparison of classification schemes clarifies what is and is not code switching for the improvement of future data gathering and analysis methods. One conclusion is that there is little justification for the educational system's prevailing stigma associated with language mixing, because linguistic borrowing is not a corruption of language but an inevitable and natural process in language contact. (MSE)