一起 may have alternate acceptable pronunciations. I looked at a number of dictionaries and see a range of results:ABC dictionary files show yi1qi3 primary but yi4qi3 as a variant (pic attached)CC-DICT dictionary files show only yi1qi3 (pics attached for mobile and web)Far East 3000 Chinese Character Dict (offline) shows only yi1qi3Oxford Concise Chinese-English Dictionary (offline) shows only yi4qi3Nciku, as phylae pointed out, shows the Collins Chinese-English, which has it as yi4qi3Am guessing either may be acceptable.
I think part of this comes down to this: Should the pinyin be written as it is said, or should be it written as-per the dictionary definition.For example, 谢谢 is pronounced xièxie. 宝宝 is pronounced bǎobao. In both cases there is simply a rule that when there are two identical characters in a row, the 2nd should be neutral. However, it is still very common very common to actually write these as xièxiè and bǎobǎo.In this case, the "standard" pinyin for 一 is yī. However, there is a rule than when 一 comes before a 3rd tone, then 一 is pronounced with a 4th tone. Thus you end up saying yìqǐ even though the you may write the pinyin as yīqǐ.I'm not an expert on this particular word. It is entirely possible that yìqǐ and yīqǐ are both legitimate pronunciations. However, I think memrise does need to choose whether its pinyin will be the "standard" version or what is actually spoken. Once this decision is made, the curators will need to start making everything consistent.
I wasn't thinking of that, is clearly the issue. Dictionaries obviously not consistent in how to handle. My own preference would be to keep tones as written for Memrise, but either way is good so long as it's consistent.Thought I read once that the official pinyin standard is that only original tone should be indicated - but don't have any reference to point to on that. And am definitely not an expert.Best solution for me would be to add an alternate pronunciation for whichever is the non-standardized variant. (This at least fixes the issue of neutral tones, which seem to vary greatly based on person, speed of speed, emphasis, etc.)
I have added in both versions here - my take is that we should accept either the standard version or what is actuallyt spoken, because both are correct, in a sense. Does that seem wrong? The other option would be to have only the version that is spoken, since that seems most logical. What do you think?