It is customary, when comparing performance of compression algorithms, to include the size of the tool needed for decompression in the compression benchmarks, s

It is customary, when comparing performance of compression algorithms, to include the size of the tool needed for decompression in the compression benchmarks, since otherwise one can simply smuggle the uncompressed data in the decompression tool. ISTM a similar principle would need to apply here: learning the "Darmok and Jalad at Tanagra" language would involve absorbing many volumes of history and mythology where for the usual sort of language a dictionary, grammar reference and maybe a book of common idioms would suffice. Whatever metric is used to compare languages for efficiency should reflect this.