“Said” for 2016 Word of the Year

I just got back from the American Association for Corpus Linguistics conference in Ames, Iowa, and I’m calling the Word of the Year: for 2016 it will be said.

You may think you know said. It’s the past participle of say. You’ve said it yourself many times. What’s so special about it?

What’s special was revealed by Jordan Smith, a graduate student at Iowa State, in his presentation on Saturday afternoon. said is becoming a determiner. It is grammaticizing.

In addition to its participial use (“once the words were said”) you’ve probably seen said used as an attributive adjective (“the said property”). It indicates that the noun it modifies refers to a person, place or thing that has been mentioned recently, with the same noun, and that the speaker/writer expects it to be active in the hearer/reader’s memory.

Attributive said is strongly associated with legal documents, as in its first recorded use in the English Parliament in 1327. The Oxford English Dictionary reports that said was used outside of legal contexts as early as 1973, in the English sitcom Steptoe and Son. In this context it was clearly a joke: a word that evoked law courts used in a lower-class colloquial context.

Jordan Smith examined uses of said in the Corpus of Contemporary American English (COCA) and found that attributive said has increasingly been used without the for several years now, and outside the legal domain. He observes that syntactic changes and increased frequency have been named by linguists like Joan Bybee as harbingers of grammaticization.

Grammaticization (also known as grammaticalization; search for both) is when an ordinary lexical item (like a noun, verb or adjective, or even a phrase) becomes a grammatical item (like a pronoun, preposition or auxiliary verb). For example, while is a noun meaning a period of time, but it was grammaticized to a conjunction indicating simultaneity. Used is an adjective meaning accustomed, as in “I was used to being lonely,” but has also become part of an auxiliary indicating habitual aspect as in “I used to be lonely.”

Jordan is suggesting that said is no longer just a verb or even an adjective, it’s our newest determiner in English. Determiners are an exclusive club of short words that modify nouns. They include articles like an and the, but also demonstratives like these and quantifiers like several.

Noun phrases without a determiner tend to refer to generic categories, as I have been doing with phrases like legal documents and grammaticization. That is clearly not what is going on with said girlfriend. Noun phrases with said refer to a specific item or group of items, in some sense even more so than noun phrases with the.

Thanks to the wireless Internet at the AACL, I began searching for of said on Twitter, and found a ton of examples. There are plenty for in said examples as well.

It’s not just happening in English. The analogous French ledit is also used outside the legal domain. Its reanalysis is a bit different, since it incorporates the article rather than replacing it. Like most noun modifiers in French it is inflected for gender and number. I haven’t found anything similar for Spanish.

In 2013 the American Dialect Society chose because as its Word of the Year. Because is already a conjunction, having grammaticized from the noun cause, but it has been reanalyzed again into a preposition, as in because science. Some theorists consider this to be a further step in grammaticization. And here is a twenty-first century prepositional phrase for you, folks: because (P) said (Det) relationship (N).

After Jordan’s presentation it struck me that said is an excellent candidate for the 2016 Word of the year. And if the ADS isn’t interested, maybe another organization like the International Cognitive Linguistics Association, can sponsor a Grammaticization of the Year.

America’s Loveliest Accents: Pittsburgh

I think it was an important move for linguists to divorce our field from aesthetics. There can be a science of taste, but science itself is not the arbiter of taste. It is not the place of linguists to judge accents or languages. Just as biologists study animals and plants that many people consider repugnant, linguists may study words and phrases that alarm or disgust people.

That said, objectivity doesn’t mean you have to like everything. Linguists have the right to our own personal tastes about languages. For example, I think German can sound very cool at times, but there are other linguists who disagree with me, and that’s okay.


The Pittsburgh dialect, with its Northern Cities-shifted vowels, is pleasant enough, but what I really love about it is its grammar. They have their own second person plural pronoun, “yinz,” from “you ones” – it’s like our “you guys,” but shorter. (Some of my neighbors have “youse,” but I never heard it in my family.) Their “an ‘at” corresponds to our “and stuff.”

The best thing about the Pittsburgh dialect, though, is that they drop the useless “to be” from certain phrases, notably “that coat needs washed.” When I visited Pittsburgh, the friendly bus driver was chatting with me on the way in from the airport. We passed a bus stop with some people sitting on a bench, and she said, “Oh, they don’t want picked up.” I wish my dialect had that! Who needs that “to be,” anyway?

Musician and comedian Mark Eddie, a native of nearby Steubenville, Ohio, plays up all these features of the Pittsburgh dialect except the “needs washed” construction in an adaptation of “Downtown,” the song made famous by Petula Clark. There are a couple of objectionable lyrics – objectifying “Shadyside chicks” and referring to transgender people as “trannies” – but beyond those you see that there’s nothing mean in his teasing; you hear a real affection for Pittsburgh and its dialect.

This is part 4 of a series where I say nice things about all of the accents that Gawker’s Dayna Evans nominated for “America’s Ugliest Accent.” Previously: New York City. Nextly: Philadelphia.

Nonstandard relative clauses

I was browsing through the Wikipedia page on relative pronouns and I found this table:

Position With explicit relative pronoun With omitted relative pronoun In formal English
Subject That’s the man [who ran away]. That’s the man [who ran away].
Direct object That’s the man [who I saw yesterday]. That’s the man [I saw yesterday]. That’s the man [whom I saw yesterday].
Indirect object That’s the man [who I gave the letter to]. That’s the man [I gave the letter to]. That’s the man [to whom I gave the letter].
Oblique That’s the man [who I was talking about]. That’s the man [I was talking about]. That’s the man [about whom I was talking].
Genitive That’s the man [whose sister I know]. That’s the man [whose sister I know].
Obj of Comp That’s the man [who I am taller than]. That’s the man [I am taller than]. That’s the man [than whom I am taller].

The “in formal English” column is a bit odd because it only contrasts to the “with explicit relative pronoun.”  I’m guessing the middle column is supposed to be “informal,” but it doesn’t include nonstandard forms that can omit a subject relative pronoun, like “There was a farmer had a dog” (observed by Arnold Zwicky among others), or those that can omit a genitive relative pronoun, like “That’s the guy’s name I was trying to remember.”