To follow up on FumbleFingers's OED quote, I note that the "to look over or through in a casual or cursory manner" definition is a relatively recent addition to the Merriam-Webster's entry for peruse. That definition first appears (in the Collegiate series) in the Tenth Collegiate Dictionary (1993). The Ninth Collegiate Dictionary (1983) has a substantially shorter entry for peruse:
1: to examine or consider with attention and in detail: STUDY 2: READ
Perhaps the most ambiguous definition of peruse in the Tenth Collegiate Dictionary is the revised form of the older definition 2 ("READ"):
2: READ; esp., to read over in an attentive or leisurely manner
This definition remains unchanged in the Eleventh Collegiate Dictionary (2003), still the most recent edition in the Collegiate Series. To me, "an attentive or leisurely manner" is a bit like "a hard-working or indolent manner"—it covers a lot of ground in two places with very little overlap between them.
Because the word can mean quite different things nowadays, I wouldn't assign a more specific meaning than "read" to any contemporary occurrence of it unless I had contextual clues to help me interpret the speaker's (or writer's) intent.
FOLLOW-UP: Various grammar and usage commentators have addressed the proper usage of peruse. For example, Eric Partridge, Usage and Abusage, Fifth Edition (1957) offers this entry:
peruse is not synonymous with 'to read', for it means to read thoroughly, read carefully, from beginning to end. One peruses a contract, one reads an (ordinary) advertisement—that is, if one does not merely glance at it.
Bryan A. Garner, Modern American Usage (2003) expresses a similar view:
peruse (= to read with great care) is pompous and stilted in business correspondence. That is, the word shouldn't be used merely as a fancy substitute for read. ... Some writers misuse the word as if it meant "to read quickly" or "scan" [examples omitted]. That slipshod extension has become common enough to be listed in some dictionaries. But since it's the opposite of the word's traditional meaning, that usage is best shunned.
Of course, both Partridge and Garner focus on how peruse should be used, not on how people actually use it. For a descriptivist view—and a critique of prescriptivist hostility toward using peruse to mean simply "read"—we can consult Webster's Dictionary of English Usage (1989). Its discussion runs for a page and a half.
Briefly, WDEU argues that peruse arose as "a literary word" that, in poetry, served as "a useful alternative to the monosyllabic read." WDEU traces hostility to the use of peruse to mean simply "read" to Frank Vizetelly, A Desk-Book of Errors in English (1906), which made the following assertion:
peruse should not be used when the simple read is meant. The former implies to read with care and attention and is almost synonymous with scan, which is to examine with critical care and in detail. A person is more apt to read than to scan or peruse the Bible.
Incidentally, scan has long since completed the bipolar metamorphosis that some people now attribute to peruse. As Garner notes, "scan is ambiguous: it may mean either (1) 'to examine carefully, scrutinize' or (2) 'to skim through, look at hurriedly.' In [American English], as it happens, sense 2 now vastly predominates—a tendency bolstered by the ubiquitous electronic scanner, which contribute to the idea of haste."
According to WDEU, "it appears that this notion of the correct use of peruse was Vizetelly's own invention. It was certainly born in disregard of dictionary definitions of the word and in apparent ignorance of the literary traditions on which those dictionary definitions were based."
WDEU then cites 21 examples (ranging in publication date from 1594 to 1968) involving peruse—some with a narrowing adverb such as thoroughly, diligently, or attentively, or (contrariwise) negligently or idly, and others with no adverbial modifiers (such as "I perused a number of public notices attached to the wall," from a 1939 book by Flann O'Brien). Then it makes its central argument:
You may have noticed by now that the plain word read can readily be substituted in any of these examples, even where the idea of "read through or over" is pretty obvious.
...
In conclusion we recommend that you reread the examples and see for yourself in how many Samuel Johnson's simple "read" definition would work perfectly well. There are likely to be only a few in which adding the adverbs used by later dictionary definers will enhance anyone's understanding of the passage.
Under the circumstances, I don't think that the Tenth Collegiate's 1993 expansion of the second definition of peruse four years after WDEU appeared from "READ" to "READ; esp., to read over in an attentive or leisurely manner" represents a disavowal by Merriam-Webster's of the generic sense "read" in favor of something along the lines of "either read in an attentive manner or read in a leisurely manner, but not simply read." Rather, I think it reflects Merriam-Webster's desire to call out the two most common narrower senses that peruse-as-"read" takes, while upholding the continued validity of the root meaning "read."
In any event, Merriam-Webster's republished WDEU as Merriam-Webster's Dictionary of English Usage in 1994 (one year after the Tenth Collegiate appeared) with its lengthy discussion of peruse unchanged.
Best Answer
Merriam-Webster's Dictionary of Synonyms, revised edition (1984) offers an extended discussion of the similar word pair retrogressive and regressive, both of which it treats as members of a group of synonyms that also includes retrograde and backward. Here is the relevant coverage:
Applying the general tenor off the distinctions between the adjective forms retrogressive and regressive to the verb forms retrogress and regress, we might say that although both involve figurative (that is, not actual physical) movement in the opposite direction from the one in which progress lies, to retrogress is to move backward (figuratively speaking) to an unspecified extent while to regress is to move backward (figuratively) in a series of incremental steps, stages, or degrees.
I must say, though, that I have no confidence that English speakers in general use these two very similar words with any such distinction in mind.
It also bears noting that retrogress is a far less common word than regress, as this Ngram chart of retrogress (blue line) versus regress (red line) versus retrogressive (green line) versus regressive (yellow line) for the period 1800–2008 indicates:
Between 1850 and 1920 or so, the frequencies of regress, regressive, and retrogressive in published writing seem to have been fairly close, after which regress and (especially) regressive took off; but retrogress has been the least common of the four forms since at least 1840, and in recent decades it has become very rare indeed.