Find solution
Is there anywhere online to check the grammar?

If Grammar is the set of structural rules, why do we still not have a software that could check grammatical mistakes accurately automatically?

  • While reading one of the most upvoted answers here on Quora, I came across this sentence : My girlfriend and I went to watch Avatar when it was released. Apparently, it was her first 3D movie experience and God was she excited. Though i did not have any problem understanding any bit of the above lines, i suspected that the highlighted part (the last four words) of the above sentence is perhaps grammatically not correct. So, I thought if i could verify this online. (This is not that i do not commit a lot of mistakes myself, in fact, a way more than an average non-native English speaker).   The top result on Google lead me to a software named Ginger http://www.gingersoftware.com/. They have an online Grammar check page so i took the above sentences, copied them six times on Google Docs, tweaked everyone of them slightly here and there as detailed below, and then checked it on this same website. To my surprise, all these sentences were declared correct by Ginger. My girlfriend and I went to watch Avatar when it was released. Apparently, it was her first 3D movie experience and God was she excited. (This is the original sentence) My girlfriend and I went to watch Avatar when it was released. Apparently it was her first 3D movie experience and God was she excited. (I removed the comma after apparently) My girlfriend and I, go to watch Avatar when it was released. Apparently, it was her first 3D movie experience and God was she excited. (See the comma placed after 'I' knowingly) My girlfriend and I went to watch Avatar when it is released. Apparently, it was her first 3D movie experience and God was she excited. (I change the tense of the 2nd part of the 1st sentence) My girlfriend and I go to watch Avatar when it was released. Apparently, it was her first 3D movie experience and God was she excited. (I change the tense of the 1st part of the 1st sentence).

  • Answer:

    Yes, there is a set of rules for how to correctly structure an English sentence, and if you are fluent in English you know most of these rules. But this set is incredibly complex and we haven't documented all of it. You could even say we don't know what the rules are. This is a bit counter-intuitive, so please bear with me. Most language processing happens in a part of the brain that is separate from the conscious mind that we all identify as "me". This part of the brain learns what is currently grammatically permissible in a language as it is learned, without your ever needing to be conscious of those rules (although being conscious of them can help your unconscious mind learn because it allows you to practice, and that "loads" your unconscious mind).  I'd like to demonstrate the difference by comparing grammar with mathematics. Imagine asking a person who is numerate to read this line and tell you whether or not it is correct: y = ( (2x - 4^2) / (x^2 - 8) )   If x = 3 then y = 7. Most numerate people can read this line and understand what each symbol means, but they won't be able to tell you whether it is correct without consciously carrying out all the sub-equations and consciously solving it. Any halfway decent programmer could write a program to check the correctness of any formula made up of parentheses and basic mathematical operators because the rules are very clear. (I realise this is not exactly analogous to grammar but the same principle applies to syntactical problems such as correct nesting of parentheses - not always easy for a person to see at first glance but very easy to express algorithmically and program into a computer). Now think of almost any line spoken by the character Yoda in the Star Wars movies. For the sake of argument, we could think of "Stopped they must be; on this all depends." Any fluent English speaker will immediately recognise that this sentence is grammatically incorrect. Ask them why, and they will pause and say the words are in the wrong order. They might offer a corrected version. Ask them to explain why one order is correct and not the other and most won't be able to give an answer anywhere near rigorous enough to build an algorithm that a computer programmer could use. Most don't even (consciously) know that English sentences follow Subject-Verb-Object sequence (the rule Yoda violates), although clearly some part of their mind knows it. This is because grammatical rules are locked up in a speaker's unconscious mind, which is able to pass only part of that information to the conscious mind. (This is what you are alluding to when you say you feel "God was she excited" is wrong but can't tell why). The Yoda excerpt is quite a simple one but there are a great many grammatically correct sentence structures, some of which are far more complex. There have been competitions offering rich prizes to the person who can correctly 'parse' the English sentence (come up with a formula that expresses all grammatically permissible structures) and those rich prizes have never been claimed. At this point in history there is simply no complete algorithm on which to base computer grammar checks, although many of the simpler, more transparent rules can be checked more or less reliably by the current generation grammar and spelling checkers. Even then, both false positives and false negatives are inevitable. This is further complicated by idiomatic expressions. These are clusters of words that are treated as a single unit of meaning (i.e. several words acting as one word). Take the expression "put up with". Why would that combination of words mean "tolerate"? Churchill once took a humorous dig at the old grammarian's rule that a sentence should not end with a preposition by twisting this idiomatic expression into "up with which I will not put". How would you even begin to write a computer algorithm to determine which of those two structures is correct? (For the record, I think "God, was she excited." is acceptable as an idiomatic construction. It seems to be a rhetorical question that was used often enough to become an idiomatic expression conveying emphasis. Otherwise, the clause "was she excited" would be grammatically incorrect as a statement (non-question) as it violates the Subject-Verb-Object rule). In addition, as has been noted in other answers, computers are at a huge disadvantage because they can't use semantic cues. , and have provided some great examples of sentences whose grammar is dependent on the semantics. If I was editing an article about dangerous marine life and came across the phrase "a man eating shark" I would correct it to "a man-eating shark". If the article was a pictorial feature about people enjoying various foods, I would do the opposite. This is a grammatical judgment but it requires semantic context. I hope my answer added something worthwhile to the excellent answers you have already received.

Trevor Best at Quora Visit the source

Was this solution helpful to you?

Other answers

I am not familiar with the grammar checker you mention, but I have used the grammar checker in Microsoft Word. I keep it turned off because it is not helpful, and is sometimes wrong. I'd say the problem is that language is just too complex. If you are talking about grammar as "structural rules," you probably mean syntax--the rules for building sentences in terms of the words' parts of speech and their arrangement into subject, predicate, object, subordinate clause, and so forth. (I would not call punctuation part of grammar.) The rules that you learned in grade school or in ESL class are not all the rules that native speakers of a language follow automatically. The rules can be used to make literally an infinite variety of grammatical sentences. Any string of 20 or so words that crops up in a conversation is likely one that has never been spoken before. And there is no theoretical limit to the length of a grammatical sentence. The software would need to break the problem up into manageable units. Keeping track of the relations between the units--like where you changed the tenses--makes the problem more difficult. (That's probably why the software didn't flag where you changed tenses--each clause, treated individually, was grammatically correct.) And considering semantics--the meaning of the words--would make things much more difficult still. Consider these two sentences. Time flies like arrows. Fruit flies like bananas. Their structure looks identical on the surface. It seems that the noun phrases on either side of the verb have been changed to make the second sentence, without altering the syntactic structure. But of course an English speaker takes semantics into account, and immediately realizes that the verb in the first sentence is "flies," and in the second sentence is "like." I think such analysis for all possible sentences would be a devilishly hard problem for software to solve. (Oh, and the last four words are grammatically correct. "God" is an interjection. "Was she excited" instead of " she was excited" is a way of adding emphasis.)

Joe Devney

Software can only best guess the components of the sentences and will easily get tangled with complex structures trying to work out what is suppose to agree with what else. Also, there are a LOT of exceptions to practically all the rules in English. I think that the progress made so far is reasonable and perhaps given time grammar checks will improved, but they're not a massively well honed tool. Part of the problem is low uptake of mechanical grammar checking because you need to know enough to know which parts to alter when you're picked up by the software, and without people clamouring for the functionality there's little incentive to improve it, I'd guess. Interestingly, personally I'd correct the second sentence to: "Apparently it was her first 3D movie experience and, God, she was excited." "God" is a sub-claus - it's an exclamation, mid sentence, that could easily be removed. That's why I'd put it in commas. If spoken, I think there would be a pause at least before God, if not after. The second half of the sentence is technically a question, albeit rhetoric, so strictly speaking it should end in a question mark rather than a full stop. However the way it would be spoken would not be representative of that part as a question. Rephrasing as I have avoids that issue.

Max Lehmann

To say that the grammar for, say, English is a set of rules, etc is not the same as saying that all of these rules have been set down formally, nor that such a set of formal rules exists which can be interpreted by a computer, no matter how powerful. As at least one other person responding to this question has suggested, there might be insufficient motivation, given the level of profit involved vs the difficulty, to produce better grammar checkers. A natural grammar is a moving target.

Bill Bell

There are many reasons, the most obvious probably being that it is unlikely to be profitable eough a venture to be worth the extensive work that would be involved, especially considering that the product would still be quite flawed. No matter how complex a program was created, it would be unable to account for such variances as context and intent. It would also tend toward failure with the most proficient users of the language, much as spell check does. And even if it took into account all the acceptable colloquialisms, regionalisms, and archaic/outdated usages, it wouldn't be able to tell you if that form of expression was appropriate for that particular usage or not.

Rebecca Billy

I am impressed with how seriously you are taking this issue.  I do not know anything about  computer programs for grammar, but Max Lehmann has given an excellent response to that part of the question, and I couldn't have written a better response myself than that of Calais Reno.  Idiomatic usage is easily understood by native speakers in every language, but is not so easy for non-native speakers, so perhaps we should think of computers as non-native speakers and leave it at that.

Mara Ferbel Goldstone

To start off with, some grammar checkers are better than others. I believe MS word in its newer versions would catch a  number of the errors you posted. That having been said, grammar is an incredibly complex set of rules often dictated by context and nuance. Consider the following sentence. I did not kick the dog. What do you think this sentence conveys? A computer would have little to no ability in discerning where the stress on this sentence is and the stress syllable can give this sentence more than six meanings. Similarly, in writing you may have cause to use unconventional grammar to convey a particular point.

Felix Zhang

Related Q & A:

Just Added Q & A:

Find solution

For every problem there is a solution! Proved by Solucija.

  • Got an issue and looking for advice?

  • Ask Solucija to search every corner of the Web for help.

  • Get workable solutions and helpful tips in a moment.

Just ask Solucija about an issue you face and immediately get a list of ready solutions, answers and tips from other Internet users. We always provide the most suitable and complete answer to your question at the top, along with a few good alternatives below.