To begin with:  they are not the same thing.

It’s true that when you’re doing research for something (and it’s hard to be a writer and not need to research things from time to time, even if what you write is contemporary literary mainstream), you want your sources to be up-to-date.  Nevertheless, there are at least three kinds of older texts that are still worth using and/or necessary to know.

First you have those essential works in a particular field that have not yet been superseded:  grammars, dictionaries, concordances, scholarly editions of primary sources, and the like.  Works of that sort tend to be difficult and time-consuming to prepare – seriously, when I contemplate the years of painstaking work it took to compile the big nineteenth-century dictionaries in fields like Old English or Old Icelandic, using only stone knives and bearskins slips of paper and filing cards, I am awed; there were giants in the earth in those days – and once a good text exists, there has to be some sort of major revolution in the field before anyone is willing to tackle the job again.^†

Then you have the groundbreaking foundational works that aren’t necessarily where the field is at any more, but that you need to be familiar with in order to understand how things got to where they are now — Sapir and Whorf in linguistics, for example, or C. S. Lewis’s The Allegory of Love in the study of medieval literature.

And then you have the works of scholarship or criticism that are literary artifacts in their own right, like Gibbon’s Decline and Fall of the Roman Empire, or, in a perverse sort of way, Robert Graves’s The White Goddess.  (Even as a bookish high school student, I saw quite clearly that as a work of anthropology or comparative religion The White Goddess was nonsense, but as a book about how Robert Graves wrote poetry, it was a fascinating document.)

Bad sources, now . . . there are also three main kinds of bad — or at any rate, considerably less than dependable – sources.  First you have books  like Holy Blood, Holy Grail, that are not just wrong in certain particulars but wrong in stereo, Sensurround, and glorious technicolor.

Then you have the surveys and popularizations and introductory texts that, however useful they may be for familiarizing a new reader with the basic outlines of a particular subject, are nevertheless bound by their very nature to be wrong in some particulars and outdated in others.

And finally, you have the sources that are hazardous to use because they’re located right where your discipline’s current controversies are taking place, and citing one of them rather than another is the equivalent of deciding which gang’s colors you’re going to wear.

^†Approach with caution, however, any book mustering copious amounts of primary-source data in the service of a Grand Theory of Everything. In my experience, Grand Theories of Everything mostly don’t work (or, as Edward Sapir put it, “all grammars leak”), and a scholar in full pursuit of a Grand Theory of Everything is in a prime position to be seduced into over-interpreting his or her data. On the other hand, they tend to collect an awful lot of it, and can be downright obsessive-compulsive about their footnotes and bibliographies.