Well there's a notorious adage, "a language is a dialect with an army and a navy".
I think academics shy away from attempting to make the distinction except when extremely obvious, and instead talk directly about quantitative measurements and feature overlaps (isogloss is a search term that may be useful here). Dialect/language lines will often have completely different shapes when you look at different distinctions in lexicon, phonetics, syntax, etc. If I had to generalize though, in a particular language "chain", linguists seem to identify an order of magnitude more separable languages than non-academics do. (Consider the cases of huge macrolanguages like "Chinese" & "Arabic", or even "Italian", whose singular labels by laypeople are pretty universally rejected.)
It doesn't help that people are generally unaware of the incredible political pressure most nations put on presenting a singular linguistic front, when the truth is much much more muddled. As a result, the common parlance distinction between dialect & language often verges on meaningless.
I think academics shy away from attempting to make the distinction except when extremely obvious, and instead talk directly about quantitative measurements and feature overlaps (isogloss is a search term that may be useful here). Dialect/language lines will often have completely different shapes when you look at different distinctions in lexicon, phonetics, syntax, etc. If I had to generalize though, in a particular language "chain", linguists seem to identify an order of magnitude more separable languages than non-academics do. (Consider the cases of huge macrolanguages like "Chinese" & "Arabic", or even "Italian", whose singular labels by laypeople are pretty universally rejected.)
It doesn't help that people are generally unaware of the incredible political pressure most nations put on presenting a singular linguistic front, when the truth is much much more muddled. As a result, the common parlance distinction between dialect & language often verges on meaningless.