There are comprehensibiilty issues with all languages, although the specific issues are language-dependent. This stems from the simple fact that language is finite and tries to map onto a much wider range of situations than the original consensus-built, common-denominator language was designed to address. Some comprehensibility issues are therefore inescapable.

In Chinese, for instance, there is a fascinating analysis of the comprehensibiity issue in Chapter 8 of Tim Clissold's book 'Mr. China'. In German, many (usually compound) words are used to express totally separate concepts, although the genesis of why these words are used so differently is easy to follow (for example - Einstellung - which literally means putting into a spot - leads to Einstellung meaning things as sundry as hiring of a worker, halting something that moves, adjustment, discontinuation, mindset, and so on, all of which mean 'putting into a spot' in some way (the word seeped into both English and French: to install / installer) .

English dealt with ambiguous words by borrowing (two-th became second, etc.) yet despite its huge lexicon it does not at all escape occasional ambiguousness (e.g., what does the sentence 'hunting lions can be dangerous' means ? Does it mean that lions when on the hunt are dangerous animals, or does it mean that people going out to hunt lions are engaging in a dangerous activity, perhaps because the cars used are unreliable in the specific hunting terrain?)

Now in your example above, when both grammar and word order are wrong, comprehensibility will still hold to some extent but I'm pretty sure that the amount of ambiguousness as to what such statements mean grows along with the incidence of syntax/grammar/vocabulary errors - i.e., say, from say 1% of all grammatically correct statements being incomprehensible, through a steadily rising % of grammatically incorrect statements against the degree of incorrectness, to at the end, from a certain degree of incorrectness, full incomprehensibility. I suppose it would be extremely hard to quantify this though.

There are comprehensibiilty issues with all languages, although the specific issues are language-dependent. This stems from the simple fact that language is finite and tries to map onto a much wider range of situations than the original consensus-built, common-denominator language was designed to address. Some comprehensibility issues are therefore inescapable.

In Chinese, for instance, there is a fascinating analysis of the comprehensibiity issue in Chapter 8 of Tim Clissold's book 'Mr. China'. In German, many (usually compound) words are used to express totally separate concepts, although the genesis of why these words are used so differently is easy to follow (for example - Einstellung - which literally means putting into a spot - leads to Einstellung meaning things as sundry as hiring of a worker, halting something that moves, adjustment, discontinuation, mindset, and so on, all of which mean 'putting into a spot' in some way (the word seeped into both English and French: to install / installer) .

English dealt with ambiguous words by borrowing (two-th became second, etc.) yet despite its huge lexicon it does not at all escape occasional ambiguousness (e.g., what does the sentence 'hunting lions can be dangerous' means ? Does it mean that lions when on the hunt are dangerous animals, or does it mean that people going out to hunt lions are engaging in a dangerous activity, perhaps because the cars used are unreliable in the specific hunting terrain?)

Now in your example above, when both grammar and word order are wrong, comprehensibility will still hold to some extent but I'm pretty sure that the amount of ambiguousness as to what such statements mean grows along with the incidence of syntax/grammar/vocabulary errors - i.e., say, from say 1% of all grammatically correct statements being incomprehensible, through a steadily rising % of grammatically incorrect statements against the degree of incorrectness, to at the end, from a certain degree of incorrectness, full incomprehensibility. I suppose it would be extremely hard to quantify this though.

Jane, that is but a reflection of the many meanings attached to a transcription into European script - for instance, "Seiko" means something like 12 very different things, but these different meanings are all written differently in Kanji

Within Japanese though, the intended meanings are mostly crystal clear (if I remember correctly , I studied Japanese pretty long ago ....)

There's something lost in translation when trying to transpose. The same thing happens from and into any language - for instance, the stark difference in meaning in English between , say, 'inevitable' and 'unavoidable' is totally lost when transposed into different European languages .....

Exactly right, Jane, but again this happens in all languages.

A famous example of that - owing to 1066 thousands of French words entered English, yet in almost all cases these words are used totally differently, and their spectrum of meaning is quite different, in English and in French.

Somebody shared a link to an Online Writing Tutor (http://virtualwritingtutor.com/, for English and French), which didn't work for me at all (showing no suggestions for my example in the original question). I used the grammar checker in Word to examine, how far it could follow the meaning of the text. Well, the first two sentences it suggested the correct use of 'he' and 'him'. The last sentence was ignored completely.

Could this be a way to find a measure for syntactical comprehensibility of texts? Get the 'best' grammar checker and analyse for how many mistakes it suggests something reasonable.

"If you say, like me, this is wrong but comprehensible, then what could be a measure to show comprehensibility here?"

In this question, I think you are not referring "comprehensibility" to a same concept. The above example of a non-English students is comprehensible to you, hence meaning your comprehensibility as a reader, but you want to design a method to test the students' comprehensibility as a text composer, correct?