I am rather spoiled by the fact that English is the, um, lingua franca, of the scientific world. My competence at foreign languages extends as far as French, to GCSE level, and German to A-level level, while my international colleagues are required to write more or less all their physics publications in English if they want them to be widely read.
Scientific English usage is not always the same as everyday usage. For example, scientists often use data as a plural, while general English usage treats it as a mass noun. There are some common constructions, though, that I think are basically incorrect, but have occurred so often in scientific papers (at least in my field) that they just slip by now and may almost be regarded as correct enough. I'm reading through a draft manuscript at the moment, and two of the most common such phrases occur:
- associated to: As in "The occurrence of the wobbling is associated to the deformation of the doobrey." This should be associated with. This seems to come about from a direct translation of the equivalent phrase in Romance languages
- allows + infinitive: As in "This super method allows to calculate terms up to fifth order". It seems a bit unfair that this logical enough construction doesn't work, but one has to say something like allows one to calculate or allows the calculation of. There are allow+infinitive usages that are okay, such as in the sense of giving permission: A allows B to do C, but not A allows to do C