There is nothing in that quote which suggests that semantics, or philosophy of language, was “belittled”, any more than observation of the nature of contemporary linguistics “belittles” comparative Indo-European linguistics. Areas of linguistics rise and fall in popularity and technical success. It is accurate to say, as the extract implies, that semantics has risen as a discipline, to the point that many graduate programs actually have multiple courses in semantics (whereas when I was a graduate student, there were none).
It should be pointed out that in the referenced time period, there were very few linguistics departments. None at Ohio State, Illinois, Washington, not even MIT. Chicago, Yale and Harvard had linguistics departments (Chicago was the first, starting in the mid-30’s, though Yale had an interdepartmental program before that). In that context, the nature of graduate study is extremely idiosyncratic, depending on the specializations of the founding faculty. The nature of Berkeley linguistics was strongly shaped by the fact that it was carried out in Anthropology. I suggest looking at the history of Berkeley linguistics page to understand the interests of linguists at the time.
The question “How did students learn ____” can be asked about many domains and time periods in linguistics: comparative IE linguistics; Semitics; psycholinguistics; formal language theory; phonological theory; stylistics; morphology; historical linguistics; discourse analysis; sociolinguistics; acoustic phonetics; lexicography; computational linguistics; forensic linguistics; second language acquisition; first language acquisition; syntax; language description. The answer is, simply, that students don’t learn every imaginable area of linguistics, and what counts as a “core” area changes over time.
The problem obliquely referred to in the quote is that given the scientific frameworks within which linguistics of the period was embedded, it was very hard reduce meaning to behavior, which meant that linguistic methods could not obviously be applied to meaning. With a change in methods, meaning became more obviously within the purvue of linguistics. Even now, linguistics is not defined as the study of everything about language. Semantics succeeded as a discipline because it was possible to show how meaning can be studied scientifically, using the methods of linguistics.
Let me add that the question is misdirected. If you want to understand the development of semantics in the history of linguistics, you ought to be asking how semantics developed in linguistics – e.g. what were those meaning properties that played a role in early linguistic theories. What view of semantics did Katz and Fodor, or Lyons, or Lakoff and McCawley, start with?
