Why Sino-Tibetan?

A correspondent asks:

Why is it that some scholars ... consider Chinese and Tibetan part of the same linguistic family even though Chinese is tonal and Tibetan non-tonal; Chinese is written using Chinese characters and Tibetan is written with a Sanskritic alphabet; Chinese uses sentence structures similar to English and Tibetan uses sentence structures similar to Sanskritic languages; and Chinese uses sentence conjugation instead of verb conjugation and Tibetan uses verb conjugation similar to that of the other Sanskritic languages?

You will be relieved to know that I am not a linguist, but it seems to me that you linguists are so smart you have discovered that dogs and chickens constitute one category of animal while cats and parrots constitute another; instead of putting the two mammals in one category and the two birds in another.

Analogies between biology and linguistics like this one are often quite useful, so let's continue in the same vein. What about whales, tuna, and buffalo? Whales and tuna both live in water, have no arms or legs, and otherwise look quite similar when compared with buffalo. On the other hand whales and buffalo are warm-blooded and give birth to live young which they nurse. Whales used to be categorized with fish, but now they are categorized with mammals--why? Because we decided that, for this kind of categorization, some traits of whales (warm-bloodedness) are more criterial than others (limblessness).

In the same way, linguists agree that the trait of shared basic vocabulary that is relatable by regular phonological correspondence is more indicative of genetic relationship than traits like tonality and sentence structure. And the writing system that a language uses is even less so. Just consider Hungarian, written with a Roman script but not genetically related to Latin or any other Indo-European language (it's Uralic). Or Urdu, written with an Arabic-based script but not related to Arabic or any other Semitic language (it's Indo-European). And when Turkish switched from an Arabic-based script to a Roman one in 1928, it did not go from being a Semitic language to being an Indo-European language; it was Turkic all along.

So we put Chinese and Tibetan in the same family because they share abundant basic vocabulary items that also are related by regular phonological correspondences. The differences in tonality, writing system, and sentence structure are readily explained by factors other than descent from a common ancestor.