Do native speakers of different languages make different mistakes when speaking in English?

For example, do native speakers of Japanese make different mistakes than native speakers of Russian when they speak in English?

Wikipedia says that pronunciation differs, for example Japanese speakers are more likely to get "R" and "L" mixed up, but I'm also interested in other things, such as grammatical errors.

  • 7
    Yes indeed they do. The very fact that there is such a thing as a recognizable "German accent" or a "Japanese accent" in spoken English means that there are language learning problems that are characteristic of one's native language. And that's just pronunciation; there are also chacteristic language-dependent syntax, semantics, and pragmatics errors. There is, in fact, a whole branch of applied linguistics -- called "contrastive linguistics" -- that's devoted to identifying and ameliorating these learning problems.
    – jlawler
    Commented Jul 24, 2014 at 15:18
  • 1
    The George Mason University Speech Accent Archive includes recordings and IPA transcriptions of native speakers from different areas, and non-native speakers from different languages, all reading the same English paragraph.
    – jlawler
    Commented Jul 25, 2014 at 2:09
  • 1
    A typical Russian mistake is "I feel badly" instead of "I feel bad"
    – Anixx
    Commented Sep 26, 2015 at 4:33

7 Answers 7


Yes, there are differences.

From my experiences as a copy editor:

  • Native speakers of Russian or Chinese (both languages don't have articles) have difficulties with articles, either leaving them out very often or overdoing them
  • Native speakers of Chinese have difficulties with tenses and with the third person singular marker on the verb
  • Native speakers of French tend to produce long-winded and complicated sentences that just don't feel right in English
  • What about Germans? Also (iii), plus too many commas?
    – Carsten S
    Commented Sep 27, 2015 at 20:10
  • @Carsten S: I am biased in this respect and probably not really good at detecting typical German errors. But too many commas is a good try, I think. Commented Sep 28, 2015 at 12:44
  • The third bullet point seems to be common among other Romance language speakers, in addition to French.
    – Pere
    Commented Nov 9, 2023 at 18:09
  • In speech, some native speakers of Mandarin mix up "he" and "she" because Mandarin does not make this distinction. Commented Nov 9, 2023 at 18:25

Recognizing accents is a relatively easy task... for people. Identifying the native language of writers is an actively researched field of Computational Linguists. So active that the field, "Native Language Identification", has its own conferences now. I paste from their website:

Native Language Identification (NLI) can be useful for a number of applications. In educational settings, NLI can be used to provide more targeted feedback to language learners about their errors. It is well known that learners of different languages make different errors depending on their L1s. A writing tutor system which can detect the native language of the learner will be able to tailor the feedback about the error and contrast it with common properties of the learner’s language. In addition, native language is often used as a feature that goes into authorship profiling, which is frequently used in forensic linguistics.

... to say nothing of helping governments identify "radicals".

The quiz, Which English, was quite popular on social media, a couple of months ago. It attempts to find out both, which dialect of English your own English is most like, and what your native language is.

  • 2
    I'm not sure what this has to do with the question. It answers how computational linguistics recognises L1 influences, not what what those influences are. Its assumptions are certainly questionable as a pedagogical tool. Commented Jul 25, 2014 at 7:48
  • @DominikLukes This is a direct answer for the question in the title. What assumptions were you referring to when you said they are questionable as a pedagogic tool?
    – prash
    Commented Jul 25, 2014 at 13:24

Obviously yes. In a lecture in Cambridge where a third of the students were fellow Germans, I once heard one of them say: "This have I not known." (This is the kind of situation the German verb fremdschämen was coined for.) Other hair-raising examples of stereotypically German English include "Can I become a steak, please" and the mythical "You can say you to me". Fortunately I learned not to say things like "I want to have an own apartment" long before I first got one in an English-speaking country. Some people are less lucky.

Perhaps more interestingly, I often catch myself forgetting the plural -s on an English noun. Although this happens to native speakers as well, I strongly suspect that in my case it's related to the fact that this particular plural ending is pretty rare in German and that the endings that we do use tend to disappear under the typical transformations that turn German words into their English cognates. (So basically, I appear to be treating English like a dialect of German to some extent, deriving a lot from German through regular transformations and then making adjustments if necessary. Sometimes the last step gets left out.)

Germans don't normally have a lot of problems with articles in English. Habit of leaving out all articles or adding the articles in exactly wrong situations is mostly reserved to the speakers of the languages that don't have the articles, such as the Russian. Confusions about singulars and plurals is most often observed in speaker of language that don't have them, like Chinese.


This is a much more complex and fraught question than it may seem. There are really two questions:

  1. Do speakers of different languages make some errors influenced by their first language? The answer here is, yes, of course. Sometimes it's called L1 interference or transfer and has been studied extensively. Contrastive linguistics (mentioned by @jlawler) thought (erroneously) that this could even be used as the basis for all language teaching. The types of errors influenced by L1 are most obvious in pronunciation but also pop up frequently in idioms, vocabulary choice and also in much of syntax. Morphology is much less susceptible to L1 interference in non-related languages. So an English person learning Russian will make similar mistakes to a Spanish person learning Russian. But a Czech person learning Russian will be more influenced by their own morphology which functions on similar principles and sounds similar.

But you can also ask:

  1. Does your native language determine absolutely the type of errors you will make in English? Then the answer is no. Not all features of native languages transfer over. And some features you would expect do not. So for example, even native speakers with rich verbal morphologies continue to make mistakes such as 'Jane walk.' even at intermediate level. So even though English verbal morphology is almost non-existent, it is still a challenge for them. Also, speakers of languages with very different syntax from English don't often transfer their word order into English in obvious ways. The interference may occur only in certain contexts and with certain constructions.

If you want more details, I recommend Rod Ellis's classic 'The Study of Second Language Acquisition'.


Learner English by michael swan enumerates all the ways non-native English speakers make errors. Grouped by language, this book is quite extensive. False cognates, word order, lack of tenses used, article usage, etc


Polish (and i think also Russian) people have often problems with differencing vowel's length (as they do not do it in their native language). Which often results in changing "sheep" into "ship" and "sheet" into... You know what.

In most other languages, even ones very closely related to Polish like Czech, vowel length is differenciated so they don't tend to make such mistakes.

Other example is the "th" (/ð/) sound, which is used in all Scandinavian languages, so naturally speakers of this languages have no problems pronouncing it, but Germans and many other nations tend to say /d/ or /z/ instead.


Yes; per p 241, Linguistics For Dummies (1 ed, 2012; by Déchaine, Burton, Vatikiotis-Bateson):

Finding transfer errors: The three ways

Linguists have figured out a pretty good way of detecting whether a language learner has transferred a property of their source language to the target language. If all three of the following patterns are found, then it’s likely that a transfer error has occurred. (This method isn’t foolproof, but it’s pretty reliable.)

Parallelism: The learner speaks the second language with patterns found in his source language. For example, speakers of Egyptian Arabic regularly insert the vowel [i] between consonants in their L2 English forms. Egyptian speakers do this because their source language doesn’t permit such consonant sequences: This is an example of how the phonotactic restrictions of the L1 source language affect the learner’s L2 language. For example, floor becomes filoor and three becomes thiree.

Homogeneity: Learners of a particular target language who also have the same source language all make the same error. For example, learners of English who have Egyptian Arabic as their source language show the same error pattern: They all break up consonant clusters with the vowel [i].

Heterogeneity: Learners with different source languages should exhibit different error patterns. For example, learners of English who have Egyptian Arabic as their source language have a different pattern of errors than learners who have Iraqi Arabic their source language. (Because these languages are both called “Arabic,” it seems like they should be the “same” language, but in fact, they are so different that a speaker of one can’t really understand a speaker of the other.) While Egyptian speakers insert the vowel [i] between consonants in their L2 English forms, Iraqi speakers insert the vowel [i] BEFORE consonant clusters. So floor becomes ifloor and three becomes ithree.

Your Answer

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

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