1. There seems to be a problem.

  2. There seems a problem.

In this type of construction, the version with to be, such as (1), is much more productive than the one without, such as (2). See this Ngram: There seems to be a,There seems a

  1. There seems to be little doubt that....

  2. There seems little doubt that....

In this similar construction with "little", however, the version with to be, such as (3), is even less productive than the one without, such as (4). See this Ngram: There seems little,There seems to be little

Would there be any logical explanation of this apparently counter-intuitive usage data?

  • Round about 2 million hits in both Google searches for "there seems to be little" and "there seems little" show currency. But productivity here surely refers to the number of different X's that will fill the slots "there seems [to be] little X ....", not directly addressed by the second two ngrams. // But I'd agree that to-be-deletion is rarely (ie not for many sensible NPs etc) available in strings of the form << There seems to be DetP/NP ... >>. Especially with NPs. << There seems reasonable ground to assume ...>> (even here, one could argue that 'reasonable' has quantifier force). May 17, 2022 at 11:28
  • 1
    There seems to be a problem is a fixed phrase and does not grammatically undergo To be-Deletion. See "To be and not to be" for details on the rule. May 17, 2022 at 13:36
  • @Greybeard Are you sure that the cited book discusses the issue raised in the OP? The linked portion doesn't seem to be particularly relevant. If I'm mistaken and you've read the book, please write up an answer even if it acknowledges that there's no logical explanation of the phenomenon.
    – JK2
    May 18, 2022 at 2:39
  • 1
    Are you sure "There seems a problem" is grammatical? I don't find it so ... May 23, 2022 at 1:08
  • 1
    @alphabet Welcome to English grammar :)
    – JK2
    Apr 29 at 23:46

1 Answer 1


The verb "to seem" is normally followed by an adjective. "It seems problematic". In the original example, the whole expression "there seems to be a problem", the "to be" is only obligatory because of the presence of "there", their combination giving the notion of existence. "There is a problem = A problem exists". So the deletion of "to be".

The uncountability (measurability) of "doubt" has nothing to do with it nor is it as strict, after all, as to forbid a sentence as "There is no doubt". "It seems doubtful". "There seems to be some/no doubt".

  • But 'there seems some doubt about' (it seems that some doubt exists about) is available as well as 'there seems to be some doubt about', so ' "to be" is only obligatory because of the presence of "there", their combination giving the notion of existence' doesn't hold water, I'm afraid, Maria. Jul 12 at 15:46
  • Edwin Ashworth: I am really not familiar with this version ("there seems some doubt"). I may not be a native English speaker, but I think that, even if available, it is not used enough as to be accepted as grammatical. Having done some more research after your comment, I only found the existential "there" combined either with "to be" or with presentational verbs.
    – Maria A.
    Jul 13 at 15:48
  • Ludwig provides examples from the New Yorker etc. But this area certainly involves divided usage; I find some of their other examples (eg 'Still, there seems some leakage') unacceptable. But Science, English Romantic Opera, Michael Palin, The Guardian, Thomson Reuters, Ofgem, the BMJ, the Science Media Centre, YouGov/Sunday Times, the Scottish Legal Aid Board, and Old Marston Parish Council are all happy to use 'There seems some doubt [about etc] ...'. Jul 13 at 19:02
  • 1
    Thank you. Facing all these, I stand corrected regarding the extent of use of the personal "seem". (And yet, if it's an equally valid variation, how come it is so often objected to?).
    – Maria A.
    Jul 14 at 12:28
  • Greenwald and Svartvik, I believe, investigated the meta-question of 'what does "acceptable" actually mean?' They found that practised linguists couldn't agree on the acceptability of some sentences. They introduced a gradience (Likert-like) model ( [C]:'60% ... acceptable' (ie say 6 out of 10 post-grad English students judge sentence 23 to be acceptable). Jul 14 at 14:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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