a) ask a different person, get a different answer. Basque gets my vote for one of those spots. If you’re talking mainstream languages then probably Japanese and Arabic.
OK, here’s my take. Modern linguistics as a discipline was mainly started by Europeans with scary beards. They applied definitions to languages which coincided with the political realities around them. This meant that for the most part, ‘languages’ nicely dovetailed with political boundaries, so French was spoken in France, Spanish in Spain and so on.
Chinese is complicated by the fact that (almost) everyone literate uses the same written language (which today is essentially a more codified form of Mandarin). So for non-Mandarin speakers (Cantonese, Hakka or whatever), this is like an French person only being able (or allowed) to write in Latin. So, the linguists come along and they see one written language, comprehensible from the Russian border to Vietnam and they think - it’s one language. Which it is, but that doesn’t apply to the spoken language(s) inside China. Traditional Chinese linguistics classifies speech by area, using the term fangyan (regional speech), but there is no distinction made for intelligibility. So Beijing has a fangyan that’s different to Taipei Mandarin fangyan, even though the two are mutually intelligible to a native speaker. But Hakka is equally classed as a fangyan, even though a Mandarin speaker could not understand it without learning the language.
This fangyan got translated into English as ‘dialect’, even though it doesn’t always accord with our idea of what a dialect is. This view fit in nicely with the old European linguist’s notion of languages defined by borders - China is one country, so it has one language. But it’s wrong. If we were to apply European standards of mutual-intelligibility-criteria to Chinese fangyan, we would come up with between seven and ten separate groups, which we could call languages. Trouble is, within those groups, there are dialects which are at best only marginally mutually intelligible, such as Northern Min and Southern Min (of which Taiwanese is part). An argument could well be made for saying that there are twenty or thirty Chinese languages.
Let’s take Taiwan’s situation as an easier problem. How many Chinese languages are spoken here? Most people would probably say three - Mandarin, Taiwanese and Hakka. The three are most definitely not mutually intelligible. OK, so linguists classify Taiwanese as part of Southern Min not a language in its own right, but effectively we’re just substituting one name for another here.
There’s a pretty good Wikipedia article on the dialect/language question that’s worth checking out.
The analysis I favour is that Chinese is a language family, and Mandarin, Wu, Hakka, Min, Cantonese, Xiang and Gan are languages within that family. Is Taiwanese a different language from the nothern Min spoken in the north of Fujian province? Tough call, but I’d say they’re dialects of the same language (Min).