I found some figures followed by "according a CNN poll" from New York Times. My instinct says it has to be "According to (then) a CNN poll" for it to be grammatical. Is it a simple mistake of the author or some other way of using "according"?

closed as off-topic by Edwin Ashworth, Drew, Glorfindel, curiousdannii, NVZ May 10 '17 at 5:22

This question appears to be off-topic. The users who voted to close gave this specific reason:

If this question can be reworded to fit the rules in the help center, please edit the question.


It is not clear to me if this is a simple production error, or a more interesting variant of the "according to" construction. In either case, I would recommend against using it yourself: "according to" is much more common and much more clearly acceptable.

Definitions of "error"

This specific case certainly might be a typo (it’s common enough for short function words like “to” to be omitted by accident).

However, there seems to be a good amount of evidence that some speakers do use the “according [noun phrase]” construction, without any preposition, with the meaning “as said by/in [noun phrase]”.

So it seems possible that the author found this construction grammatical; in that case, it would not be a simple “production error”.

It could still be called an “error”, but only if we use this word to mean something else that is probably harder to define. For example, someone who maintains that this construction is an “error” even if it was used intentionally by the author might mean any of the following:

  • “I find it grammatically unacceptable”
  • “Most English speakers would find it grammatically unacceptable”
  • “It is rarely found in standard written English, and a good publisher would only let it be published through oversight”

This is where evidence comes in, since we can see if evidence supports or contradicts any of these statements.

There is some evidence suggesting it is not a production error, but I haven't found a ton of it

The Oxford Dictionaries entry for "according" cited by other answers is evidence that the construction is used often enough to have been noticed by lexicographers. This indicates that there are probably some people to whom it is grammatically acceptable, even if these people are a minority. (The statement “usually according to” indicates that at minimum, most people prefer the version of this construction with “to”.)

It is also possible to find a number of examples of the construction without to from a Google Books search, which is often a helpful approximate indicator of English usage in published text.

  • The diffusion rate of electric appliances has continued to be advancing at a steady pace, according a survey by the Japan Electric Machine Industry Association.

    The Oriental Economist, Volume 34, Issues 663-674 (1966)

  • According a trade source, the value of contracts concluded between Japan and China at the Kwangchow Trade Fair showed a decline.

    Technocrat, Volume 7 (1974)

  • Sandista forces killed 136 insurgents and lost 28 soldiers in northeastern Nicaragua during the last month, according a spokesman for the Nicaraguan Defense Ministry.

    Latin American index, Volumes 10-13 (1982)

  • According to Interior the mine fires began July 1960 at the Centralia dump and have been monitored for carbon dioxide and carbon monoxide emissions since 1980. In 1983 the underground fires caused the ground at one location in the town to collapse, according a department spokesman...

  • (another page) Excluded from coverage by the regulation are small containers of waste produced by laboratories and placed in drums, according a technical assistant at the state’s Pollution Control Board.

    Environment Reporter, Volume 16, Issues 1-26 (1985)

    I think these last two examples are especially interesting as they occur in the same publication, and the first one is near an instance of “according to”. This could possibly mean that the author thinks of the a constructions as having different meanings or uses, although it could just be a case of free variation. Alternatively, it could support the hypothesis that the omission of "to" is an unintentional production error or typo.

Evidence suggesting it is just a common production error

Here are two examples I found that definitely supported the "production error"/"typo" hypothesis:

Each of these has one single example of "according the", and many more examples of "according to the" (15 for the first, 21 for the second).


I wasn't aware of this usage before. The example from your question, "according a CNN poll", is grammatically unacceptable for me, and apparently also for Drew (an American English speaker) and Glasseyed (who seems to be a Canadian resident). Edwin Ashworth (a UK English speaker), says he would not use it, although he doesn't find it totally unacceptable. I haven't seen a comment from anyone saying that they do use the construction.

Since the construction "according to [noun phrase]" is more common and as far as I know is entirely acceptable for all speakers, that is the form that I would recommend using.

I don't think that there is enough data, or that I am good enough at analyzing it, to be able to say much about who uses "according [noun phrase]"; and in fact, I'm not sure even after going through corpus results that it is a genuine grammmatical variant of the "according to [noun phrase] construction": it still seems possible to me that in the examples that can be found, it is either a production error, or perhaps a non-native formation in some cases. I haven't found any document where "according [NP]" is used consistently, or even at a near-equal rate to "according to [NP]".

  • Thank you very much for such a through answer and research! – N.R. in Seoul May 9 '17 at 3:06
  • Not 'unacceptable'. I've mentioned the Quirk-Svartvik five-point scale for 'acceptability' before now, and can't place the prepositional usage at 'totally unacceptable' after reading the ODO article. But I wouldn't use it. – Edwin Ashworth May 9 '17 at 12:37
  • This is a good example of a question addressing a grey area in English. And of a balanced response. Rather than the all-too-common 'No it isn't / Yes it is' subjective/unsupported answers. – Edwin Ashworth May 9 '17 at 13:23
  • 1
    A good, thorough answer & exposition, sumelic. Here's what I consider a shorter version: 'according to' is never wrong in the context quoted, it needs no explanation, and won't raise questions in the minds of people unfamiliar with common English usage. They won't have to consult online dictionaries, or the CNN manual of style (if there is one). – Glasseyed May 9 '17 at 21:33

The adverbial expression is usually followed by "to", but there are usage examples (see Ngram) without the preposition "to":

According (usually according to)

As stated by or in:

  • ‘the outlook for investors is not bright, according to financial experts’


  • The Ngram uses are minimal; in many Ngram uses without to, there's a scan error or an old Biblical use. – Xanne May 9 '17 at 3:47

No, the to is common but not required, as definitions will tell you:

(usu. according to) as stated by or in:
we have the world's most expensive public transport, according a recent survey.
Oxford Dictionaries

  • 1
    Saying "... according a recent survey" would seem to many readers/listeners to imply that "according" is a verb and the survey is being given some "accord" -- being granted agreement. – Hot Licks May 8 '17 at 23:18
  • 1
    What @HotLicks said. To my AmE ears, leaving out to sounds very wrong in this context. I wonder if that isn't maybe a BrE thing? (Just a guess.) – Drew May 8 '17 at 23:29
  • 1
    @Drew It sounds either incorrect or American to my ears. But ODO is a respected dictionary, with respected corpus researchers. – Edwin Ashworth May 8 '17 at 23:35
  • Laurel, I'd be interested in knowing, does this construction sound acceptable to you as an individual without the "to"? – sumelic May 9 '17 at 1:55

Your instinct is correct. "According a" is either an oversight or a sign of…well, I wouldn't care to speculate.

  • A sign that you're prepared to accept that research-based dictionaries are almost always a better guide than one person's limited experience? – Edwin Ashworth May 8 '17 at 23:33
  • No, a sign that some people are prepared to accept the current OED's notorious willingness to embrace usages that previous editions would have dismissed as colloquial or illiterate. If you are prepared to accept it, by all means feel free. – Glasseyed May 8 '17 at 23:52
  • 2
    Oxford Dictionaries is not same thing as the OED, and I have never heard before of either having a reputation for accepting "usages that previous editions would have dismissed as colloquial or illiterate". (I have heard people say such things about, e.g. Webster's Third New International Dictionary and current Merriam-Webster dictionaries.) – sumelic May 9 '17 at 0:25
  • I've read several such comments (and similar comments about the Third New International, of course), but regret to say I can't find them. And since I only own the first OED, I can't point to examples. By the way, thanks for pointing out that the ODO is not the same thing as the OED: I'm not in the habit of consulting online dictionaries, and in my ignorance assumed they were one & the same. :-) – Glasseyed May 11 '17 at 18:55

Not the answer you're looking for? Browse other questions tagged or ask your own question.