May 23, 2013, 12:01 am

Speech Recognition vs. Language Processing

I have stressed that we are still waiting for natural language processing (NLP). One thing that might lead you to believe otherwise is that some companies run systems that enable you to hold a conversation with a machine. But that doesn’t involve NLP, i.e. syntactic and semantic analysis of sentences. It involves automatic speech recognition (ASR), which is very different.

ASR systems deal with words and phrases rather as the song “Rawhide” recommends for cattle: “Don’t try to understand ‘em; just rope, ride, and brand ‘em.”

Labeling noise bursts is the goal, not linguistically based understanding.

Current automatic-speech-regcognition systems cannot reliably identify arbitrary sentences from continuous speech input. This is partly because such subtle contrasts are involved. The ear of an expert native speaker can detect subtle differences between detect wholly nitrate or holy…

May 22, 2013, 12:01 am

More Important(ly)

By Anne Curzan

An author publishes a book containing a sentence beginning with “Even more strikingly, …” His mother points this out to him as “a glaring error,” feeling strongly that the sentence’s main clause should not be introduced with the adverb strikingly but rather with the adjective striking. The son disagrees. The question lands in my e-mail box for arbitration.

The mother is far from alone in her objection, but the objectors’ case is not a strong one.

As I wrote to the mother, it is fine to use either the adverb or the adjective in this sentence: “Even more strikingly, …” or “Even more striking, …” Strikingly functions as a sentence adverb here (modified by even more to make an adverb phrase), expressing the writer’s attitude or stance toward the assertion contained in the main clause. The writer believes that what follows is more striking than what came…

May 21, 2013, 12:01 am

Prospecting for Antedates

By Allan Metcalf

Remember 1849? Those were the great days of the California Gold Rush. Hundreds of thousands dropped everything to grab gold from the foothills near Sutter’s Fort. In that heady time, you didn’t need lots of equipment—perhaps just a pan to sift riverbed gravel for nuggets.

Well, it’s 1849 all over again. Not in gold mining, which now generally requires sophisticated technology, but in etymology, the study of word origins. Vast new fields of data have been opened and made accessible, so it’s easier than ever to find an earlier instance of a word or phrase not yet recorded in any dictionary.

Last week I gave an example of antedating, the Yale librarian Fred Shapiro’s discovery of an 1886 hot dog in a Nashville newspaper, some six years earlier than any previously discovered use of that now-familiar name for a sausage in a bun.

Years ago, before the Internet, that kind…

May 20, 2013, 12:01 am

Dueling Titles

By Lucy Ferriss

Hundreds of readers opened their New York Times Book Review recently to see a review of a novel that had already been reviewed in April . . . no, wait. That earlier book was Life After Life by the terrific British novelist Kate Atkinson. This book is Life After Life by the terrific American novelist Jill McCorkle. A galumphing typo by the compiler of the table of contents at NYTBR? Nope. There’s the review, glowing about McCorkle’s book much as the reviewer of Atkinson’s book had glowed a mere two weeks earlier.

You cannot copyright a title, and good thing too. Otherwise, the dozen iterations of Forever that have appeared in print in the last two years alone (romance, fantasy, werewolves, YA—name your own genre) would have to resort to the thesaurus for Evermore, Ever and Anon, Till Hell Freezes Over, Semper Eadem. But although McCorkle’s and Atkinson’s publishers are…

May 17, 2013, 12:01 am

Hot Dog!

By Allan Metcalf

News flash in the etymological world: Two new antedatings of hot dog!

In the etymological world, prospecting for earlier instances of a word is like prospecting for gold in the geological world. You look in the online Oxford English Dictionary for the earliest known date of a word and then go data mining in the archives of old publications for something earlier.

One of the leading prospectors is Fred Shapiro of the Yale University Library. He announced his findings in the first instance earlier this year on the American Dialect Society’s discussion list, ADS-L. That got the attention of the assay office, a.k.a. Comments on Etymology, a paper and ink journal I’ve written about before. It’s published by Gerald Cohen at the Missouri University of Science and Technology, and it’s the first draft of etymological history. The news about hot dog came in the recently arrived Vol. 4…

May 16, 2013, 12:01 am

Spit That Image Out

By Lucy Ferriss

Quickly, now, without checking any dictionaries or usage guides: which of the following expressions is original, standard usage?

Once and awhile
Set and stone
Try and get
Spit and image
All and all
Hand and hand
Tongue and cheek

I’ve run into all of these recently, mostly in student papers, but also in published work. So many of our habitual expressions have lost their connection to the original meaning that students—and sometimes professional writers—set them down as they sound without regard to whatever sense they might make. Given the aural similarity of and, in, and -ing, it’s no surprise that malapropisms like in this day in age crop up—for how often do we actually think about something being common in this day and also in this age, or era? And why would we?

If we think carefully or have background knowledge, of course, we can and do make some sense…

May 15, 2013, 12:01 am

Never Abolish the To-Die-For Sentence

By Ben Yagoda

Word came—via Twitter, Tumblr, I don’t remember, something that starts with a t—that The New Yorker has been featuring on its Web site the five best sentences of the week. That was good to hear, as I collect great sentences, the way some people collect beach glass, small statues of turtles, or perceived insults.

I was disappointed to find, however, that “Backblogged: Our Five Favorite Sentences of the Week” consists of sentences from a rather small subset of published work, The New Yorker itself. No one admires The New Yorker more than I do. However, I judge a magazine, even The New Yorker, to be too small a sample to yield each week five sentences worthy of collecting: that is to say, sentences which you cannot think of a way to improve and which might have a chance of living on when the immediate circumstances of their publication are long forgotten. Here, for example, is…

May 14, 2013, 12:01 am

In Art No One Can Hear You Scream

By William Germano

“The Scream” (1893), by Edvard Munch; mug from Yizzam.com

Last summer I had the chance to see two versions of a work by Norway’s greatest artist, Edvard Munch. If you go to Oslo you can see it in one version at the National Gallery (no photos, please) and another in the lovely Munch Museum (cameras welcome). In recent years each of them has been stolen and recovered. A third version of the work has been on view this year at MoMA.

Of course, you know the picture—it’s an icon of modernity’s anxiety. The figure has been thought to depict a psychiatric patient, or even a mummy.

Munch himself spoke of his synesthetic response to the world, where ideas and words presented themselves in chromatic terms. His recollection of a particular urban moment inspired him to create the various versions of a visual arrangement that would…

May 13, 2013, 12:01 am

Keyword Search, Plus a Little Magic

By Geoffrey Pullum

I promised last week that I would discuss three developments that turned almost-useless language-connected technological capabilities into something seriously useful. The one I want to introduce first was introduced by Google toward the end of the 1990s, and it changed our whole lives, largely eliminating the need for having full sentences parsed and translated into database query language.

The hunch that the founders of Google bet on was that simple keyword search could be made vastly more useful by taking the entire set of pages containing all of the list of search words and not just returning it as the result but rather ranking its members by influentiality and showing the most influential first. What a page contains is not the only relevant thing about it: As with any academic publication, who values it and refers to it is also important. And that is (at least to some extent)…

May 10, 2013, 12:01 am

Singular ‘They’: a Footnote

By Anne Curzan

Over the past two decades, the use of they as a singular generic pronoun has been defended often and eloquently by linguists in various venues, including here on Lingua Franca. Geoff Pullum has written about the topic twice in the past year and a half: “Dogma vs. Evidence,” and “We Do Not Seek to Rule.” Pullum and others have written about the use of they with a singular antecedent extensively on Language Log. The Lousy Linguist, in yet another defense of the construction, provides a useful list of Language Log posts on the topic, through 2008.

I myself have taken part in the effort to defend singular generic they on numerous occasions. I happily climb on my soapbox about this construction because (a) they is singular in common usage, so it doesn’t make sense to call it “ungrammatical” in the descriptive sense (it is completely meaningful to both speakers and…