"I love ambiguity more than most people" is of course ambiguous, since it could mean "I love ambiguity more than most people (love ambiguity)" or "I love ambiguity more than (I love) most people." And in the case of some linguists, both of those propositions may have positive truth values.

(For more on the ambiguity of "comparative ellipsis," see Jean Mark Gawron, "Comparatives, Superlatives, and Resolution," Linguistics and Philosophy 18:333-380, 1995.)

Not since Bill Clinton challenged the definition of "is" has so much hinged on a very short word.

John Edwards appears to basing much of his defense, which begins today in a North Carolina courtroom, on the legal interpretation of the word "the." […]

The statute governing illegal receipt of campaign contributions "means any gift, subscription, loan, advance, or deposit of money… for the purpose of influencing any election for federal office."

The words "the purpose" suggests that in order for a conviction, the sole reason for the money would have to be to finance a presidential campaign.

Edwards' legal team has argued … that his main reason for hiding Hunter was to keep her secret from his wife, Elizabeth.

Prosecutors, however, are arguing the law should be interpreted to mean "a purpose," meaning use of the donations does not have to be solely for a political campaign.

Once during the anti-Vietnam war protests, when many of us at UCLA were restructuring courses to not be “business as usual”, I found this lovely book on language and the law: Bryant, Margaret M. 1962. English in the law courts; the part that articles, prepositions, and conjunctions play in legal decisions. New York,: F. Ungar. I used it in my semantics class for the rest of the semester, and realized what a fertile field language and the law could be (and it  indeed blossomed in subsequent decades.) It had a whole chapter on “the” (other chapters included "and", "or", and "of". I was in heaven!)  It’s been more than 40 years since I read it, but the main thing I remember is that for every position any semanticist or logician has taken on any of the little logical words in the language, some judicial decision can be cited as precedent for that position. (Larry Solan's first book, The Language of Judges (1993, University of Chicago Press) made the same point about potential parsing ambiguities — all possible ambiguity resolution principles are attested in judicial precedents.) And certainly the “uniqueness” condition on "the" is open to dispute.

There is now a whole research area devoted to “weak definites” — the term was coined, I believe, by Massimo Poesio in a 1994 paper, though the relevant phenomena had been observed much earlier. Sentences like “The kids had been writing with indelible markers on the wall of their grandmother’s living room” occur frequently, even though living rooms have four walls, and the meaning is the same as would be expressed by the (less natural sounding) “a wall”. A frequently cited sort of “weak definite” occurs in “My father was reading the newspaper and didn’t look up when Dick came in.” “Reading the newspaper” can just name an activity, much like “newspaper-reading”. Not every occurrence of a definite article can be so interpreted; one of the challenges is to explain why we can have a weak definite in “read the newspaper” but not in “read the book.” There will be a conference (not the first by any means) on weak definites in Florianopolis, Brazil in August. Perhaps the Edwards example will by then be a stock example.

I trust the prosecution team will have some linguists on call and/or some linguistically savvy lawyers. If the claim that "the purpose" in "for the purpose of" can only mean "the sole purpose" is the best defense Edwards has, I suspect he’s in trouble.

[Thanks to Jonathan Lighter via Ben Zimmer for the pointer to the Yahoo/Good Morning America article.]

THERE is an ongoing debate in this country about the rich: who they are, what their social role may be, whether they are good or bad. Well, consider the following. A recent study found that 10 percent of people who work on Wall Street are “clinical psychopaths,” exhibiting a lack of interest in and empathy for others and an “unparalleled capacity for lying, fabrication, and manipulation.” (The proportion at large is 1 percent.) […]

The only thing that puzzles me about these claims is that anyone would find them surprising.

The only thing that puzzles *me* about such claims is that they spread so far in reputable publications, over such a long period of time, despite being complete fabrications.

Deresiewicz (or the NYT editors) do provide a link to the 10-percent-of-Wall-Streeters-are-psychopaths "study", which goes to an article in The Week, "Why is Wall Street full of psychopaths?", 3/1/2012:

One out of 10 Wall Street employees is a clinical psychopath, estimates Sherree DeCovny in CFA Magazine [pdf for purchase], compared with one out of 100 people in the general population. That statistic is "shocking," says Sam Ro at Business Insider, and, while it doesn't mean Wall Street is crawling with ax-wielding serial killers (see American Psycho), the extreme character traits outlined in DeCovny's report are certainly prevalent — and often admired — in the industry.

The article in The Week, in turn, gives us a link to what seems to be the original source of the statistic —  Sherree DeCovny, "The Financial Psychopath Next Door", CFA Magazine, March/April 2012.

When I read the Deresiewicz study, my science-journalism spider sense started tingling.  The fact that the "study" link went to another news report kicked the tingle up a notch; and my skepticism crossed the threshold to near certainty at the step to CFA Magazine, a "practice-based, professional member magazine" which

…contracts with professional journalists to write most articles published in CFA Magazine. Accordingly, stories are approached from a journalistic perspective, providing varying viewpoints that are representative of the CFA Institute global membership and that offer real-life, professional applications through a balance of theory and practice.

So the question was, did I care enough about this example of the science-journalism telephone game to pay CFA Magazine a few dollars for a copy of the article? Before deciding, I thought I'd look around for a copy posted somewhere on the web; and that simple web search quickly revealed that an expert job of debunking had already been done: John Grohol, "Untrue: 1 out of Every 10 Wall Street Employees is a Psychopath", Psych Central 3/6/2012:

[I]n trying to research where this statistic came from, I stumbled upon a symptom of what’s wrong with a lot of journalism today.  I can summarize the problem in one word — laziness. Many (most?) journalists nowadays take “experts” words for whatever claims they make, without ever bothering to check them out independently.

Alexander Eichler, a “business reporter” at The Huffington Post, started this news cycle by making the claim in his article, “One Out Of Every Ten Wall Street Employees Is A Psychopath, Say Researchers:”

One out of every 10 Wall Street employees is likely a clinical psychopath, writes journalist Sherree DeCovny in an upcoming issue of the trade publication CFA Magazine (subscription required). In the general population the rate is closer to one percent.

Eichler isn’t suggesting 1 out of 10 Wall Street employees is a psychopath — he’s just passing along something he read in another magazine (note, it’s a magazine, like People, not a scientific journal). Eichler’s point of his blog entry is simply to regurgitate what DeCovny (2012) wrote in her article. Here’s what DoCovny, a freelancer, actually wrote:

Studies conducted by Canadian forensic psychologist Robert Hare indicate that about 1 percent of the general population can be categorized as psychopathic, but the prevalence rate in the financial services industry is 10 percent. And Christopher Bayer believes, based on his experience, that the rate is higher.

When DeCovny was contacted about the statistic, she replied:

Christopher Bayer, a psychologist I interviewed for the article, told me about Hare’s study, so he should be able to point you in the right direction. Christopher provides therapy to Wall Street professionals. He’s also finishing up a book on this topic.

It’s great that Christopher Bayer is a therapist who treats Wall Street professionals. However, I could find no research he’s authored in this topic area. So while his opinion is duly noted, it really isn’t in the same league as empirical scientific data. The two should never be confused.

Hare, on the other hand, is a famous researcher who has made a career in studying psychopaths, has published dozens of scientific studies on the topic, and developed the preeminent checklist that is used in most psychopathy research. […]

Hare did indeed co-author a paper that examined “corporate psychopathy,” with colleagues Paul Babiak and Craig Neumann (2010). It did not look at the financial services industry specifically. The research used a sample that consisted of 203 corporate professionals from 7 different companies, selected by their companies to participate in management development programs from all areas of industry.

I did what any journalist writing about a famous researcher should do before saying he said something that seems a little “out there” — I contacted Hare to ask him about this data. Here’s his response to the claim that 1 in 10 (10 percent) of financial industry employees is a “psychopath:”

I don’t know who threw out the 10% but it certainly it did not come from me or my colleagues.

The article to which you refer describes a sample of “203 corporate professionals selected by their companies to participate in management development programs.” The sample was not randomly selected or necessarily representative of managers or executives, or of the corporations in which they work.

The approximately 4% who had a PCL-R score high enough for a research description as psychopathic cannot be be generalized to the larger population of managers and executives, or to CEOs and the “financial services industry.”

So to be crystal clear here, one out of every ten wall street employees is NOT a psychopath. At least not according to any actual scientific research. DeCovny took a professional’s word (Bayer’s) that this is what the research showed; and she had no reason to doubt him. But she also didn’t verify the information for herself (like I did), or bother contacting Hare to ensure the data being attributed to him was correct. (We could not reach Christopher Bayer in time to comment on the discrepancy between what he told DeCovny, and what Hare actually researched.)

Just to clarify a small point of quantifier scope: By saying that "one out of every ten wall street employees is NOT a psychopath", Dr. Grohol does not mean that 10% are non-psychopaths, implying that 90% of them ARE psychopaths; but rather than it is not the case that 10% of them are (or at least have been shown to be) psychopaths.

Now, I'm not a reflexive defender of Wall Streeters. For all I know, 10% — or 90% — of them really are psychopaths. But I do believe that public policy debates ought to be based as much as possible on facts, and not on made-up numbers that have been given a pseudo-scientific status by a multi-step journalistic "telephone game".

Adding to the documentation of this particular stream of nonsense in the media sewer, Dr. Grohol gives a list of other sightings:

• The Huffington Post: One Out Of Every Ten Wall Street Employees Is A Psychopath, Say Researchers
• Business Insider: The Shocking Statistic About Psychopaths On Wall Street
• Current TV: Researchers: One in 10 Wall St. employees is a psychopath
• Nonprofit Quarterly (NPQ): Wall Street Bonuses Larger Than Most Nonprofit Executive Directors’ Salaries
• Decline of the Empire: Psychopaths On Wall Street

Regular readers may recall some of the dozens of other examples that we've picked up on over the years, widely-reproduced stories telling us that email and texting lower IQ twice as much as smoking marijuana does; that Twitter and Facebook numb our sense of morality and make us indifferent to human suffering; that women talk three times as much as men do, while men think of sex every 52 seconds on average; and so on. In an earlier post, I concluded that

… "scientific studies" like these have taken over the place that bible stories used to occupy. It's only fundamentalists like me who worry about whether they're true. For most people, it's only important that they're morally instructive.

What would the [editors responsible for promulgating these stories] say, if presented with evidence that they've been peddling falsehoods? I imagine that their reaction would be roughly like that of an Episcopalian Sunday-school teacher, confronted with evidence from DNA phylogeny that the animals of the world could not possibly have gone through the genetic bottleneck required by the story of Noah's ark. I mean, lighten up, man, it's just a story.

For those interested in following up further, the study referenced above is Paul Babiak, Craig Neumann, and Robert Hare,  "Corporate psychopathy: Talking the walk", Behavioral Sciences & the Law, 2010. I'll have a bit more to say about it in a later post.

The number of people who use "whom" and "who" wrongly is appalling. The problem is a difficult one and it is complicated by the importance of tone, or taste. Take the common expression, "Whom are you, anyways?" That is of course, strictly speaking, correct - and yet how formal, how stilted! The usage to be preferred in ordinary speech and writing is "Who are you, anyways?" "Whom" should be used in the nominative case only when a note of dignity or austerity is desired. For example, if a writer is dealing with a meeting of, say, the British Cabinet, it would be better to have the Premier greet a new arrival, such as an under-secretary, with a "Whom are you, anyways?" rather than a "Who are you, anyways?" - always granted that the Premier is sincerely unaware of the man's identity. To address a person one knows by a "Whom are you?" is a mark either of incredible lapse of memory or inexcusable arrogance. "How are you?" is a much kindlier salutation.

The Buried Whom, as it is called, forms a special problem. That is where the word occurs deep in a sentence. For a ready example, take the common expression: "He did not know whether he knew her or not because he had not heard whom the other had said she was until too late to see her." The simplest way out of this is to abandon the "whom" altogether and substitute "where" (a reading of the sentence that way will show how much better it is). Unfortunately, it is only in rare cases that "where" can be used in place of "whom." Nothing could be more flagrantly bad, for instance, than to say "Where are you?" in demanding a person's identity. The only conceivable answer is "Here I am," which would give no hint at all as to whom the person was. Thus the conversation, or piece of writing, would, from being built upon a false foundation, fall of its own weight.

A common rule for determining whether "who" or "whom" is right is to substitute "she" for "who," and "her" for "whom," and see which sounds the better. Take the sentence, "He met a woman who they said was an actress." Now if "who" is correct then "she" can be used in its place. Let us try it. "He met a woman she they said was an actress." That instantly rings false. It can't be right. Hence the proper usage is "whom."

In certain cases grammatical correctness must often be subordinated to a consideration of taste. For instance, suppose that the same person had met a man whom they said was a street cleaner. The word "whom" is too austere to use in connection with a lowly worker, like a street-cleaner, and its use in this form is known as False Administration or Pathetic Fallacy.

You might say: "There is, then, no hard and fast rule?" ("was then" would be better, since "then" refers to what is past). You might better say (or have said): "There was then (or is now) no hard and fast rule?" Only this, that it is better to use "whom" when in doubt, and even better to re-word the statement, and leave out all the relative pronouns, except ad, ante, con, in , inter, ob, post, prae, pro, sub, and super.

According to Thomas Fensch, The Man Who Was Walter Mitty: The Life and Work of James Thurber, 2001 (p. 145), Thurber published nine of these pieces on usage: "Who and Whom"; "Which"; "The Split Infinitive"; "Only and One"; "Whether"; "The Subjunctive Mood"; "Exclamation Points and Colons"; "The Perfect Infinitive"; and "Adverbian Advice".  Unfortunately, The New Yorker's on-line archive apparently only indexes "abstracts", so that the only way to find the other eight is to browse every issue in the 1928-1930 time period. This is interesting but time-consuming — so I invited commenters with subscriptions (and thus archive access) to offer more detailed citations.

Update — the full set of nine is reprinted here, as they appeared in The Owl in the Attic (1931). And in the comments, Dave Lull cites a bibliography that gives the New Yorker issues  in which they were originally published.

(More "Whom Humor" … and fans of grammatical humor shouldn't miss Geoff Pullum's "The coming death of whom: Photo evidence", 9/10/2004.)

[For those who didn't learn Latin in an old-fashioned way, I'll note that "ad, ante, con, in, inter, ob, post, prae, pro, sub, and super" is a list that schoolchildren used to be required to memorize, because verbs compounded with these prepositions generally govern the dative case. Thurber has left off "… and sometimes circum", but he seems to have managed to make his way in the world nevertheless.]

I was catching up on my stack of New York Times magazines, and I came across a mini-article in their "One-Page Magazine" feature from January 15. I couldn't find it on their website, but here's the entire content:

How do you write that in Mandarin?
by Mireille Silcoff
Chinese characters comprise the world's oldest in-use writing system, but Chinese kids are forgetting how to get it on paper. The new term tibiwangzi ("take pen, forget characters") encapsulates the issue: nobody takes pen anymore. They type or text, often using Romanization. The China Youth Daily Social Survey Center says 4 percent of respondents are "already living without handwriting."

A fuller treatment of this subject may be found in an article entitled "In information era, handwriting growing obsolete in China" that appeared in the April 16, 2010 People's Daily.

The main point is certainly consistent with the character amnesia that Victor Mair has written about before; but I was struck by the claim in the first clause. It's not necessarily false, but it's not obviously true either. In fact, I would say a strong competitor for that title is the (Semitic) alphabet. (In what follows, I use alphabet somewhat broadly for a system in which symbols stand for single sounds, which are just consonants in the typical Semitic variety; some prefer the more specific term abjad.) How might one evaluate the claim?

1. Chinese writing is attested from around 1250 in the oracle bone inscriptions (all dates here are BCE, and approximate). But the near-certain origin of the alphabet, in Proto-Sinaitic and Proto-Canaanite writing, is no later than 1700. (We won't count the much older Egyptian system that inspired this alphabet.)

2. Chinese writing has always represented the same language, or at least the ancestor of the same language, although (pre-) Old Chinese is very different indeed from modern Mandarin. But the early alphabetic scripts represent a Northwest Semitic language, which would be the ancestor of languages like Hebrew and Aramaic, or very closely related to their common ancestor. (We can't really say that the oracle bone inscriptions encode a dialect that is precisely the ancestor of Middle Chinese, anyway.) But this criterion will exclude Arabic, since it's not part of the Northwest branch of the family; its ancestor was not written until much later than Chinese.

3. Though the claim simply says "in use", one might narrow this to a language with a large continuous community of native speakers. That would disqualify Hebrew, but the alphabet did remain in vigorous continuous use for Hebrew despite a lack of native speakers. Since Aramaic has continued to be spoken in pockets of the Middle East, however, this language would still qualify for continuous native speakers.

5. If we want to emphasize the visual sameness of the system, note that Chinese writing was standardized around 220, but the various local Semitic alphabets are established by around 900, and the Aramaic letters used to write Hebrew take their essential current form by 500 or so.

4. Chinese characters, from their beginnings, are structurally the same as the modern characters, most importantly in the existence of composite characters with a semantic and phonetic element (such as 洗 xǐ "wash", xiǎn "a surname" with the phonetic 先 xiān plus the water semantic or radical). But the actual characters with this structure are often quite different, in both appearance and composition (choice of components). For the alphabet, the basic structure – one symbol represents one sound – is in place in 1700, even if the appearance changes and some specific characters are dropped or added in later versions.

The main argument one could make against the structure of the later Hebrew or Aramaic alphabet being the "same" as Proto-Canaanite, comparable to the structural sameness of the oracle bone inscriptions and modern Chinese writing, is the representation of vowels. In the second millennium, as far as we know, the symbols only represent consonants, and all vowels are unrepresented. From the early first millennium, most of the implementations of the alphabet, including Hebrew, begin to indicate more and more vowels by means of secondary uses for consonantal letters, e.g. y for /i/. Since these letters retained their consonantal values, however, it's not clear that these secondary uses constitute a basic change in the system to match that of the Greek alphabet, with dedicated vowel letters.

Floating around the claim about Chinese, I think, is a view of the language, nicely phrased by Bob Ramsey, "as if it were one unchanging, monolithic entity across space and time". This applies to the written form as well. If one converts an Old Chinese inscription into modern equivalent characters (not always easy to do), then a modern speaker can "read" it with modern pronunciations. This gives the false impression that it's exactly the same writing system. A similarly cursory comparison of modern Hebrew with Proto-Canaanite could also efface some of the differences, but the fact that the alphabet represents sounds alone will make it harder to ignore the changes that have occurred over the centuries.

The advantage of Chinese for this purpose is that the syllabic components to the complex characters can be "read" in modern Mandarin instead of Old Chinese, without attending to the massive changes in pronunciation that occurred between the two. For example, the words represented by those characters 洗 and 先 have been reconstructed by Baxter and Sagart as *sˤərʔ and *sˤər for Old Chinese (a language that didn't even have tones yet). If we overlook these differences in the linguistic content of the symbols for Chinese, then we probably should do the same for the alphabet, and in that case the apparent greater antiquity of the Chinese writing system has less to do with its actual history than the way people look at the language and the script.

John Rickford wrote that "the patterns of variation and change … are regular rather than random, governed by unconscious, language-internal rules and restrictions" — but Ms. Acocella took this defense of "vernaculars that are commonly regarded as lacking rules", from a scholar known for his defense of "Ebonics", as a stalwart affirmation of prescriptive standards.

Steven Pinker tried to explain how false beliefs about standard usage, like No Split Verbs or No Final Prepositions, can become widespread — but Ms. Acocella took this attempt to distinguish between true and false beliefs, from the author of a popular book on "Words and Rules: The Ingredients of Language", as promoting the idea that "there are no rules", other than the false "old wives' tales" he debunked.

If you've read Acocella's review, you will have noticed something else about this hallucinated debate: she's really angry about it. In particular, she doesn't care for Hallucinated Steve Pinker at all.

You can see this animus in her gratuitous comment about his choice of terminology:

There are no rules, he declares. Or they’re there, but they’re just old wives’ tales—“bubbe-meises,” as he puts it, in Yiddish, presumably to show us what a regular fellow he is.

Pinker's own account of this choice:

I call them bubbe meises, Yiddish for “grandmother’s tales,” in tribute to the late language columnist William Safire, who called himself a language maven, Yiddish for “expert.”

Acocella's anger emerges again in excoriating the AHD's editors for publishing Pinker:

Most important is that the editors tried to pull descriptivists over to their side. In the most recent edition, the fifth, they have not one but two introductory essays explaining their book’s philosophy. […] For the editors of the A.H.D. to publish Pinker’s essay alongside Rickford’s is outright self-contradiction. For them to publish it at all is cowardice, in service of avoiding a charge of élitism.

But Pinker is the chair of the dictionary's Usage Panel, and so his essay, explaining "Usage in the American Heritage Dictionary", was included ex officio, not recruited as part of some sinister plot to placate descriptivists. (And the idea that Steve Pinker might have been recruited to protect John Rickford from "a charge of elitism" is truly hilarious.)

Acocella's ire at Evil Descriptivists emerges even more strongly in her discussion of the book she's reviewing, Henry Hitchings' The Language Wars: A History of Proper English.  Since Hitchings debunks false claims about allegedly proper usage, he must be one of those "anything goes" guys — but wait, he admits that there are also real linguistic regularities! In her final two paragraphs, she combines this hallucinated contradiction with a common complaint about the hypocrisy of writers who use standard English in defending vernacular usage:

… the A.H.D.’s run for cover is not as striking as the bending over of certain descriptivists, notably Hitchings. Having written chapter after chapter attacking the rules, he decides, at the end, that maybe he doesn’t mind them after all: “There are rules, which are really mental mechanisms that carry out operations to combine words into meaningful arrangements.” We should learn them. He has. He thinks that the “who”/“whom” distinction may be on its way out. Funny, how we never see any confusion over these pronouns in his book, which is written in largely impeccable English.

No surprise here. Hitchings went to Oxford and wrote a doctoral dissertation on Samuel Johnson. He has completed three books on language. He knows how to talk the talk, but, as for walking the walk, he’d rather take the Rolls. You can walk, though.

If a student so badly misunderstood simple ideas expressed in clearly-written texts, the usual response would be to bemoan the decline of critical reading skills in kids today. But Ms. Acocella is no student — and her misreading of works on the norms of usage has, I suspect, been  part of her magazine's culture for more than fifty years.

Consider this passage from E.B. White's letter to J.G. Case, his editor at Macmillan for The Elements of Style, dated 17 December 1958 (emphasis added):

I was saddened by your letter — the flagging spirit, the moistened finger in the wind, the examination of entrails, and the fear of little men. I don't know whether Macmillan is running scared or not, but I do know that this book is the work of a dead precisionist and a half-dead disciple of his, and that it has got to stay that way. I have been sympathetic all along with your qualms about "The Elements of Style," but I know that I cannot, and will-shall not, attempt to adjust the unadjustable Mr. Strunk to the modern liberal of the English Department, the anything-goes fellow. Your letter expresses contempt for this fellow, but on the other hand you seem to want his vote. […] In your letter you are asking me to soften up just a bit, in the hope of picking up some support from the Happiness Boys, or, as you call them, the descriptivists. (I can write you an essay on like-as, and maybe that is the answer to all this; but softness is not.) […]

All this leads inevitably to like-as, different than, and the others. I will let them lay for the moment, sufficient unto this day being the etc. My single purpose is to be faithful to Strunk as of 1958, reliable, holding the line, and maybe even selling some copies to English Departments that collect oddities and curios. To me no cause is lost, no level the right level, no smooth ride as valuable as a rough ride, no like interchangeable with as, and no ball game anything but chaotic if it lacks a mound, a box, bases, and foul lines. That's what Strunk was about, that's what I am about, and that (I hope) is what the book is about. Any attempt to tamper with this prickly design will get nobody nowhere fast.

P.S. When I said, above, that Macmillan would have to take me in my bare skin, I really meant my bare as.

(For more of the letter, see "No smooth ride is as valuable as a rough ride", 2/19/2005.)

Key elements of Acocella's 2012 screed are present in White's 1958 letter: in particular, the view that publishers are conspiring behind the scenes to placate hypocritical liberal descriptivists, "little men" who believe that "anything goes".

And a brief examination of E.B. White's "bare as" may help explain his anger — and Acocella's. There's an excellent discussion of this issue in the entry for like, as, as if in Merriam-Webster's Dictionary of English Usage. I invite you to read the whole thing, which I will summarize briefly here.

In the 1950s, shortly before E.B. White's letter to his editor, an advertising slogan ("Winston tastes good, like a cigarette should") created a storm of controversy in the popular press about the use of conjunctive like. In the New Yorker, the editors Viewed With Alarm this "obnoxious and ubiquitous couplet", with its "pesky 'like'". Strunk (and Strunk & White) condemned this "illiterate" usage, which "lately … has been taken up by the literate, … who use it as though they were slumming".

There was a problem with this picture, as some of those Happiness Boys noted: it got the facts wrong. To quote MWDEU:

… information published in the Middle English Dictionary shows that like by itself was used as a conjunction as long ago as like as was — from the late 14th century. […] Chaucer used it in about 1385 to introduce a full clause in The Complaint of Mars.

Conjunctive like was also used by Shakespeare and others around 1600, but it becomes rare during the 17th and 18th centuries. However,

In the 19th and 20th centuries conjunctive like becomes much more common. Jespersen 1909-49 (vol 5) tells us that "example abound" and lists them from Keats, Emily Bronte, Thackeray, George Eliot, Dickens, Kipling, Bennett, Gissing, Wells, Shaw, Maugham, and others. So we must conclude that Strunk & White's relegation of conjunctive like to misuse by the illiterate is uninformed.

This is the kind of "attempt to adjust the unadjustable Mr. Strunk" that E.B. White was so upset about. And when deeply-felt beliefs and allegiances come into stark conflict with the facts, an emotional response is normal.

Sometimes, people who are faced with such conflicts can adjust their beliefs and allegiances to accord with reality. But others respond with denial, willful misunderstanding, and conspiracy theories.  You can see that sort of response in climate-change denialists — and it seems that the staff of the New Yorker have been English usage denialists for more than half a century.

Is there going to be a language log comment on the article "The English Wars"  in the current issue of the New Yorker?  I find it completely shocking to see that an author who purports to be writing about prescriptivism vs. descriptivism has so little understanding of the subject, and that the editor (presuming that such a position still exists at the New Yorker) would not catch the absurd claim that John Rickford is a prescriptivist.

The passage that shocked Philipp is the following, discussing the history of the American Heritage Dictionary:

Nowadays, everyone is moving to the center. The big fight produced some useful discussions of linguistic history, including Guy Deutscher’s “The Unfolding of Language” (2005). These books, by demonstrating how language changes all the time, brought about some concessions on the part of the prescriptivists, notably the makers of the A.H.D.’s later editions. First, the editors changed the makeup of their advisory panel. (The original hundred advisers were not dead white men, but most of them were white men, and the average age was sixty-eight.) Some definitions were made more relativist.

Most important is that the editors tried to pull descriptivists over to their side. In the most recent edition, the fifth, they have not one but two introductory essays explaining their book’s philosophy. One is by John R. Rickford, a distinguished professor of linguistics and humanities at Stanford. Rickford tells us that “language learning and use would be virtually impossible without systematic rules and restrictions; this generalization applies to all varieties of language, including vernaculars.” That’s prescriptivism—no doubt about it. But turn the page and you get another essay, by the cognitive psychologist Steven Pinker. He tells us more or less the opposite. There are no rules, he declares. Or they’re there, but they’re just old wives’ tales—“bubbe-meises,” as he puts it, in Yiddish, presumably to show us what a regular fellow he is. And he attaches clear political meaning to this situation. People who insist on following supposed rules are effectively “derogating those who don’t keep the faith, much like the crowds who denounced witches, class enemies, and communists out of fear that they would be denounced first.” So prescriptivists are witch-hunters, Red-baiters. For the editors of the A.H.D. to publish Pinker’s essay alongside Rickford’s is outright self-contradiction. For them to publish it at all is cowardice, in service of avoiding a charge of élitism.

This passage is indeed deeply confused. Its author, Joan Acocella, is the New Yorker's dance critic, and either the topic was not felt to be important enough to merit elementary editorial supervision, or there is no one at the magazine with any competence in the area involved.

Steve Pinker's essay wasn't exactly stuck in at the last minute by craven AHD editors — Pinker is the chair of the AHD's usage panel, for goodness' sake, and his piece describes the panel's approach to questions of usage.  The previous usage panel chair was Geoff Nunberg, whose qualifications included a widely-cited 1983 article in the Atlantic Magazine, "The Decline of Grammar", which attacked such language mavens as John Simon and Edwin Newman as "shrill" and "dogmatic", and was attacked in turn by Mark Halpern in a 1997 Atlantic article "The War that Never Ends".

And in fact, there's no contradiction or even tension between  John Rickford's essay ("Variation and Change in Our Living Language") and Steve Pinker's ("Usage in the American Heritage Dictionary").

John Rickford's central point is that vernacular forms of language are not degraded or mistaken approximations to an ideal standard, but rather living systems with their own internal logic.  For hundreds of years, linguists have described this internal logic of language in terms of laws or rules, even — and perhaps especially — when the patterns of usage are changing. Thus Rickford observes that

… the patterns of variation and change … are regular rather than random, governed by unconscious, language-internal rules and restrictions that can often be appreciated only when we assemble large numbers of examples and study these quantitatively. People tend to think of rules and grammar as covering only the small set of items about which we receive overt instruction […]. But in fact we are unconscious of most of the language regularities and restrictions that we follow every day. […] Language learning and use would be virtually impossible without systematic rules and restrictions; this generalization applies to all varieties of language, including vernaculars.

Many of the entries for which we have provided Our Living Language notes in this volume are similarly subject to systematic rules that their speakers follow regularly, if unconsciously, even though these words and constructions come from vernaculars that are commonly regarded as lacking rules.

There is no contradiction at all between this perspective and the one that leads to Steve Pinker's observation about what he calls bubbe-meises:

In the absence of publicized regulations like traffic laws, the elevation of shared knowledge to common knowledge can be unpredictable, even chaotic. Outlandish fashions, surprise bestsellers, dark-horse candidates, currency hyperinflations, and asset bubbles and crashes are all cases in which people behave according to the way they expect other people to expect other people to expect other people to behave. The craving for common knowledge can even lead to a false consensus, in which everyone is convinced that everyone believes something, and believes that everyone else believes that they believe it, but in fact no one actually believes it. […]

The maddening paradox of false consensus has long afflicted lexicographers and grammarians. The problem goes by various names— folklore, fetishes, superstitions, bugaboos, and hobgoblins—but I call them bubbe meises, Yiddish for “grandmother’s tales,” in tribute to the late language columnist William Safire, who called himself a language maven, Yiddish for “expert.” A grammatical bubbe meise is a rule of usage that everyone obeys because they think everyone else thinks it should be obeyed, but that no one can justify because the rule does not, in fact, exist. The most notorious grammatical bubbe meise is the prohibition against split verbs, where an adverb comes between an infinitive marker like to, or an auxiliary like will, and a main verb. According to this superstition, Captain Kirk made an error when he declared that the five-year mission of the starship Enterprise was to boldly go where no man has gone before; it should have been to go boldly. Likewise, Dolly Parton should not have declared I will always love you, but I always will love you or I will love you always.

Steve's suggestion that "everyone obeys"  superstitions like No Split Verbs is a bit misleading, as his own examples indicate. But his goal in this passage is to explain how belief in such false principles can become so widespread, not to deny that any genuine linguistic rules exist. He's published dozens of scientific articles about linguistic rules, and even a popular book "Words and Rules: The Ingredients of Language".

Such rules — the rules that John Rickford writes about — are the emergent regularities of living language, including vernacular varieties of language. In contrast, Steve Pinker's "bubbe meises" (he doesn't use the word rule in describing them) are invented stipulations about allegedly proper usage, promulgated by explicit instruction. This is close to Friedrich Hayek's distinction between "rules which have by a process of selection been evolved", in what he calls a "grown order", and what he calls a "made order", governed by principles that are "thought of as having been deliberately constructed by somebody, or at least owing whatever perfection they possessed to such design". Rickford and Pinker don't disagree in any fundamental way about the nature of these two kinds of rules or principles, any more than Hayek disagreed with himself about his two kinds of social order.

As Hayek observed, this debate encompasses "all institutions of culture … [m]orals, religion and law, language and writing, money and the market". The New Yorker as an institution has traditionally been on the "made order" side of this debate in most areas, and especially so in matters of language. But that's no excuse for publishing such a confused and badly-informed review.

Update — more here.

(Full disclosure: I'm a member of the American Heritage Dictionary's Usage Panel.)

[For discussions of other aspects of Acocella's article, see Jan Freeman, "Descriptivists as hypocrites (again)", Throw Grammar from the Train 5/8/2012; John McIntyre, "Cheap Shot", You Don't Say, 5/82012; and "Ignorant Blathering at the New Yorker", Language Hat 5/11/2012. For a historical and philosophical perspective on the 'scriptivisms, see Geoff Nunberg's Language Log post, "The politics of 'prescriptivism'", 11/20/2011. And for more of my own views on the matter, see e.g. "Amid this vague uncertainty, who walks safe?", 2/23/2007, or "Menand on linguistic morality", 10/22/2008. ]

I think I hear Obama use 'extraordinary' and 'extraordinarily' a lot (an 'extraordinary' amount). Is there a way to check this in your data?

I responded

In 127 speech transcripts, in a total of 110,100 words, Obama uses extraordinary 17 times and extraordinarily once. That's a combined rate of 1000000*18/110100 = 163 per million words. In the 425-million-word COCA corpus, extraordinary occurs 13,360 times and extraordinarily 2,701 times, for a combined rate of 1000000*(13360+2701)/425000000 = 38 per million words.

So relative to the language at large, he (or his speech-writers) do use extraordinary a lot.

How this compares to political oratory from other sources is a different question.

Eugene followed up:

[W]ouldn't a president talk about extraordinary things from time to time?

So I thought I'd look into this a bit more, over lunch.

I downloaded from the web site of the UCSB American Presidency Project everything in the  "Oral: Address - Saturday Radio" category. This yielded 1345 transcripts of presidential Saturday radio messages, as follows:

And what are the extraordinary numbers for these five presidents?

So except for Reagan, the others do use extraordinary at about twice the background rate; but Obama's rate is about four times the background. (The rates for Bush 1 are hard to interpret, because his total word count in this particular category is so small.)

Some other aspects of the frequency counts are slightly intresting.

George W. Bush used enemy and enemies a lot; Ronald Reagan was fond of weakness; Bill Clinton was big on smoking:

Update — As long as we're comparing these collections of presidential radio addresses, let's take a look at the rates of first-person singular pronoun usage:

What do these numbers mean?

Nothing at all, except in the context of statements like this:

George Will: If you struck from Barack Obama’s vocabulary the first-person singular pronoun, he would fall silent, which would be a mercy to us and a service to him, actually.

I recognize that Mr. Will is speaking hyperbolically. But the clear meaning of his hyperbole is that Barack Obama uses first-person singular pronouns excessively often; and in that context, this otherwise-meaningless comparison of rates also acquires a meaning, namely that George Will is careless with the truth.

For links to an excessively large number of posts making similar points, see here.

I have come to realize that there are a lot of examples on the web of the type "not want to not X" that seem to say the opposite of what they mean. Here are a few:

She failed to give the patient CPR and turned an ambulance away in the mistaken belief that the elderly woman’s had said she did not want not to be resuscitated. (Cambridge, UK, newspaper article)

If a guest does not want not to be disturbed they need only to place the 'Do Not Disturb" sign on the door and their wishes will be respected. (Florida motel)

In the first case, the mistaken belief was that the elderly woman did not want to be resuscitated. In the second case it should say "If a guest does not want to be disturbed …"

I first thought that these were isolated errors but there are so many of them that they cannot not all be errors. You find this phenomenon with "wish," "desire," and even "expect:"

If you do not wish not to abide by these terms and conditions, please do not use this site and leave immediately.

Nor will we waste your time or ours emailing you information you do not wish not to receive.

I do not desire not to overwhelm the inquirer with great amounts of reading, nor overwhelm myself with great amounts of writing on this subject.

If the girl remained unkissed, she could not expect not to marry the following year.

Do you know if something has been written about this phenomenon? It doesn't look like negative concord.

As Arnold observed in response, this appears to be a new type of misnegation, of the overnegation variety, not previously catalogued. And Larry Horn noted

I can only add that one feature that often leads to (or at least accompanies) hypernegation, including the cases illustrating what I call "Triplex negatio confundit" and "Quadruplex negatio far'blondiat" (in my old 1991 CLS paper on "logical" double negation), is the lexical incorporation of (at least) one of the extra negatives.  That property is present in some of the examples below ("She failed to give the patient CPR and turned an ambulance away in the mistaken belief that…"; "Nor will we waste your time…", "If the girl remained unkissed,…", and arguably also "they need only…"), so that may be a factor even when the incorporated negatives are in a different clause.  In the third example, perhaps the "If not p then not q" form may have thrown a monkey wrench into the mix; I wonder if the hypernegation would have arisen had the consequent been simply "please leave immediately".  In the remaining example, there's a "nor" anticipated in the later clause.  I'm not claiming there's an inevitable relation between the processing of some negations (which is known to put a strain on both production and comprehension, back to work by P. C. Wason, Herb Clark, and others) and the difficulty of keeping track of others, but there does seem to be a correlation.

Larry's 1991 CLS paper is discussed here — or you can read his 2009 BLS paper here.

Update — There has been some discussion in the comments about whether these examples are instances of editing errors, perhaps facilitated by interaction with word processing programs. FWIW, we can certainly find cases of this kind that were written well before the era of computerized word processing, such as this passage from the address of Walker D. Hines to the National Lumber Manufacturers Association, published in Lumber World Review, April 23, 1919:

Broadly speaking, the policy of the Railroad Administration is that the advisors on each railroad will continue to make their purchases just as they did before the war and on the same general basis with, however, some limitations to prevent the exercise of undue pressure for a considerably lower price. The Railroad Administration could get no ultimate advantage by prusing such a short-sighted policy. It has no such desire nor has it a desire to bring about prices for itself which are below the prices of of other purchases of commodities in substantial volume. It does not want not to be the beneficiary of special treartment which will result in putting a burden on the rest of the public.

The shot of the column got plenty of attention when Craig Silverman of Poynter's Regret the Error posted it, speculating that it was the result of a spellcheck error (what we call a cupertino in these parts). The Huffington Post also picked it up, and pretty soon it was getting spread all over Twitter and Facebook.

On Twitter, however, Alex Johnson, a reporter at MSNBC.com and ex-Post staffer, told Silverman he was "almost certain" that this was merely a joke by his former colleague de Moraes — along the lines of her nickname for American Idol's Ryan Seacrest, "Seabiscuit." And soon enough, de Moraes cleared up the matter in an update to the online version of the column. It was indeed a gag, and she credited the nickname to a participant in one of her weekly online chats about TV. (She also tipped her hat to the originator of the "frumious Bandersnatch," Lewis Carroll.) Her regular readers probably got the joke without the need for an extra wink, but that was lost on those who merely saw a howler in the pages of a distinguished newspaper.

Moral of the story: before indulging in viral schadenfreude, consider whether the joke might be on you.

And then takes it in a somewhat unexpected direction:

Such things are certainly all over the internet, often in the writings of people who are clearly well past their student days:

Would of like to of come, but I'll still be in West Palm Beach.
Just would of like to of seen the World of Tanks and the Museum get a little more exposure.
I would of like to of seen a playmaker in there at least somewhere in midfield,

But the fact is, traditional spelling aside, "I would of like to of seen it" is exactly how I pronounce the phrase myself.

Some of the examples that turn up on the web underline the mainstream character of this pattern.

Consider this review on urbanspoon.com of a restaurant in Portland OR:

This is a very quaint little restaraunt with incredibly great food and a dessert menu that would be the envy of any top pastry shop in Paris! Great place for a special lunch or business lunch.
Everything on menu is made fresh and to order.
The selections are unique items, with bold flavors and quality ingredients.
I got the Croque Monsieur sandwich which was Black Forrest Ham, Gruyere Cheese, on toasted Parmesan crusted bread. The sandwich was served very warm, with cheese perfectly melted. It was delightful!
Instead of Fries I got the House Salad. It was very large with hazel nuts and mixed greens. Dressing was light and would of like to of had just a bit more.
Dining partner got the Mediterranean Vegetable Sandwhich. Served on ciabatta bread and it was bursting with flavor.
Fondue appetizer was almost a meal in itself! A rich, creamy crock pot of bold tasting cheese fondue accompanied by apple slices, asparagus and homemade bread slices. The bread's texture was just full of little nooks and crannies that just sopped up and captured the fondue cheese! Yummy!
Have only eaten here for lunch, but would love to try dinner sometime.
And if you go, save room for desert! I didn't and wish I had!

Or this interview with "British action star Mem Ferda" (where the choice of spelling is the transcriber's, of course):

Q: Are there any roles you would of liked to of played?

A: I would of like to of played Stanley Kowalski in A Streetcar Named Desire, Travis Bickle in Taxi Driver, and Jake La Motta in Raging Bull. In my later years, I would like to be Shakespeare’s King Lear at The Globe.

Or this comment at the Cleveland Plain Dealer site, on a photo showing the city council president holding a coat over his head:

Yes, yet another one of those special candid 'keeper' photos by the outstanding PD photographer Marvin Fong. I'm still chuckling whenever I see it because so many times many of us would of like to of put a muzzle and/or a bag over PIG Big Daddy's mouth/head; so he finally obliged us! However, sure wish that he would of taken the stand as it would of been a howl to see when eventually he couldn't contain himself when angered and would go off ranting all his vulgar obscenities!

If you struck from Barack Obama’s vocabulary the first-person singular pronoun, he would fall silent, which would be a mercy to us and a service to him, actually.

I call his assertion a lie, rather than using the technical term "bullshit", because it seems that Mr. Will is trying to convince us of something that is patently untrue, rather than just trying to puff himself up without any particular concern for whether what he says is true or false.

This is not an accidental or casual remark — George Will has been beating this particular drum since June of 2009. For a list of earlier posts on this topic, including more counting of first-person pronouns than you could possibly want to read about, see "A meme in hibernation", 3/31/2012. (Hint: In comparable kinds of material, Barack Obama's rate of using first-person-singular pronouns is low compared to the usage rates of other recent presidents.)

Adding one more source to the list, I give below the counts of first-person-singular pronouns from the presidential radio addresses on file at The American Presidency Project at UCSB (technically, everything in the "Oral: Address - Saturday Radio" category):

I recognize that when Mr. WIll says that "if you struck from Barack Obama’s vocabulary the first-person singular pronoun, he would fall silent",  he's speaking hyperbolically. But the clear meaning of his hyperbole is that Barack Obama uses first-person singular pronouns excessively often; and in that context, this otherwise-meaningless comparison of rates also acquires a meaning, namely that George Will is careless with the truth.

Update — more from Fred Vultee here. And following up a question in the comments below, I'd like to suggest that you consider this quotation from George Will:

Look, self-absorption is part of the occupational hazard of politics, and it’s also part of the job description of being president. All that said, try to imagine Dwight Eisenhower talking about D-Day saying, ‘I did this. I decided this. I did this and then I did that.’ It’s inconceivable.

And then take a look at these two quotations from Dwight D. Eisenhower (in an interview with Walter Cronkite in a CBS program "D-Day Plus 20 Years", quoted in Loudon Wainwright, "D Day Reminder of the Best Ike", Life Magazine, 6/16/1964):

I thought it was just the best of a bad bargain. I possibly sat silently just reviewing these things, maybe, I'd say 35 or 45 seconds. . . . Actually, I think after 30-45 seconds of ssomething like that, I just got up and said, "Okay, we'll go," and this room was emptied in two seconds."

"Because if it did fail," he told commentator Walter Cronkite cheerfully, "I was going into oblivion anyway, so I might as well take full responsibility."

Let me repeat again the questions that I asked in an earlier post:

There are two interesting questions here, it seems to me. The first one is why George F. Will is so struck by rates of first-person usage, on the part of Barack and Michelle Obama, that are significantly lower than has been typical of recent presidents and first ladies on similar occasions. The second question is how many pundits and talking heads will follow his brainless lead this time around. […]

Now that I think of it, there's another significant question here as well. How in the world did our culture  award major-pundit status to someone whose writings are as empirically and spiritually empty as those of George F. Will?

[Hat tip to David Yamanishi.]

Three items since then:

BZ, 4/16/12: The first “asshole” in the Times? (link)

AZBlog, 4/29/12: Annals of French taboo avoidance (link)

and today: AZBlog, 5/7/12: Reporting the profane (link)

Is there any discussion anywhere of the multiple tense-marking (if that's what it is) in constructions like "We would have liked to have stayed longer" (as opposed to just "We would have liked to stay longer")?  And is it just my impression, or has this become more common?

For what it's worth, there's a very clear discussion of what the difference theoretically could be here.

Web search turns  up the original lyrics to Elton John and Bernie Taupin's song Candle in the Wind, which includes the line "I would have liked to have known you". You would think some Telegraph reader might have made the connection between the song's popularity and the decline of the English language, but if that happened I can't find any evidence of it.

Henry Watson Fowler is probably not what Bob means by "some Telegraph reader"; but Fowler 1926 says

After past conditionals such as should have liked, would have been possible, would been the first to, the present infinitive is (almost invariably) the right form, but the perfect often intrudes, & this time without the compensation noted in 1, the implication of non-fulfilment being inherent in the governing verb itself.  So:– If my point had not been this, I should not have endeavoured to have shown the connexion. / Jim Scudamore would have been the first man to have acknowledged the anomaly./ Peggy would have liked to have shown her turban & bird of paradise at the ball./ The Labour members opened their eyes wide, & except for a capital levy it is doubtful whether they would have dared to have gone further.

(For more on the "implication of non-fulfilment", see the end of the post.)

Many other usage guides follow Fowler, often more emphatically, in viewing the double perfect marking in such cases as illogical and wrong. Thus Garner 2009:

would have liked. This phrase should invariably be followed by a present-tense infinitive–hence would have liked to go, not *would have liked to have gone, *would have liked to have read. The erroneous phrasings are very common–e.g.:

• "One would have liked to have been [read would have like to be] present at the meeting in which the introduction of this equipment was ratified." Giles Smith, "'Replay' Ends Dispute over Hurst's Goal," Daily Telegraph, 16 Aug. 1997, at 21.
• "Clapp said he would have like to have seen [read would have liked to see] more teams involved in postseason play," Richard Olbert, "expanded Playoffs Rejected," Ariz. Republic, 29 Aug. 1997, at C12.

Such phrasings, erroneous or not, have certainly been common for a long time. Here's Disraeli:

The world goes on with its aching hearts and its smiling faces, and very often, when a year has revolved, the world finds out there was no sufficient cause for the sororws or the smiles. There is too much unnecessary anxiety in the world, which is apt too hastily to calculate the consequences of any unforeseen event, quite forgetting that, acute as it is in observation, the world, where the future is concerned, is generally wrong. The Duchess would have liked to have buried herself in the shades of Brentham, but Lady Corisande, who deported herself as if there were no care at Crecy House except that occasioned by her brother's rash engagement, was of opinion that 'Mamma would only brood  over this vexation in the country,' and that it would be much better not to anticipate the close of the waning season. So the Duchess and her lovely daughter were seen everywhere where they ought to be seen, and appeared the pictures of serenity and satisfaction.

And Bram Stoker:

It seemed to me that we were simply going over and over the same ground again; and so I took note of some salient point, and found that this was so. I would have liked to have asked the driver what this all meant, but I really feared to do so, for I thought that, placed as I was, any protest would have had no effect in case there had been an intention to delay.

And Thackeray:

The dirty fist in the young surgeon's pocket was obliged to un-double itself, and come out of its ambush disarmed. The poor fellow himself felt, as he laid it in Pen's hand, how hot his own was, and how black - it left black marks on Pen's glove; he saw them, - he would have liked to have clenched it again and dashed it into the other's good-humoured face; and have seen, there upon that ground, with Fanny, with all England looking on, which was the best man - he, Sam Huxter of Bartholomew's, or that grinning dandy.

And Oscar Wilde:

The exaggerated folly of the threat, the passionate gesture that accompanied it, the mad melodramatic words, made life seem more vivid to her. She was familiar with the atmosphere. She breathed more freely, and for the first time for many months she really admired her son. She would have liked to have continued the scene on the same emotional scale, but he cut her short.

And Ernest Hemingway:

Frances was a little drunk and would have liked to have kept it up but the coffee came, and Lavigne with the liqueurs, and after that we all went out and started for Braddocks's dancing club.

Here's some data from COHA, which suggests that such phrasings may have been becoming more rather than less common, despite the best efforts of Fowler, Garner, and their ilk:

Responding to Bob's note, Geoff Pullum gave a grammatical and logical analysis of why Fowler (though not Garner) might be right:

Having thought about this before, I'm inclined to say that there is no "double marking".  I think these sentences:

[1]     I would like to know Marilyn.
[2]     I would like to have known Marilyn.
[3]     I would have liked to know Marilyn.
[4]     I would have liked to have known Marilyn.

are all four grammatical, with different meanings.  Let t₀, t₁, and t₂ be three time points in chronological order, t₂ being now. I think the meanings are:

[1′]    Speaking at t₂ I would like at t₂ to know Marilyn at t₂.
[2′]    Speaking at t₂ I would like at t₂ to be in a state of having known Marilyn at t₁.
[3′]    Speaking at t₂ I would have liked at t₁ to know Marilyn at t₁.
[4′]    Speaking at t₂ I would have liked at t₁ to have known Marilyn at some earlier point t₀.

If there are mistakes out there, I think they involve people saying one that isn't the one they really meant; but some of the truth-conditional differences are subtle, so not only would it be hard to convince someone they had made an error, it would be hard even to convince yourself.

Bob Ladd responded:

Right - I agree that the meanings are all potentially distinct, and in fact that's exactly what it says at the useful link I included in my earlier message.

But my empirical observation was simply that (to use your numerical references) people often use construction [4] when they want to convey meaning [3′].  Sometimes it's hard to tell, of course, but in general meaning [4′] doesn't make much sense.  […]

It's for that reason that I think we're dealing with double marking, at least in the way construction [4] is normally used.  Calling these mistakes is dangerously close to suggesting that somebody who says "I didn't see nothing" really means "I saw something".

For what it's worth, I seem to be among those who prefer what Bob calls "double marking" — "I would have liked to have known you" seems better to me than "I would have liked to know you", in the meaning that Geoff identifies as [3′]. As Fowler suggests, the double marking feels like a way of emphasizing the irrealis character of the embedded clause:

1. After past tenses of hope, fear, expect, & the like, the perfect infinitive is used, incorrectly indeed & unnecessarily, but so often & with so useful an implication that it may well be counted idiomatic. That implication is that the thing hoped &c. did not in fact come to pass, & the economy of conveying this without a separate sentence compensates for lack of logical precision.  So :– Philosophy began to congratulate herself upon such a proselyte from the world of business, & hoped to have extended her power under the auspices of such a leader./ It was the duty of that publisher to have rebutted a statement which he knew to be a calumny./ I was going to have asked, when . . .

The section on The Perfect Infinitive in The King's English (1908) is even more sympathetic to perfect infinitives as irrealis markers:

This has its right and its wrong uses. The right are obvious, and can be left along. Even of the wrong some are serviceable, if not strictly logical. I hoped to have succeeded, for instance, means I hoped to succeed, but I did not succeed, and has the advantage of it in brevity; it is an idiom that it would be a pity to sacrifice on the altar of Reason.

In this context, "strictly logical" can be glossed as "interpreted according to the illogical premise that each morphological category has an a single and invariant meaning".  As often, the Fowlers' linguistic intuitions are better than their linguistic theories.