Take that same encyclopedia, but imagine it being read and transcribed onto the same string of words with punctuation and all (no difference so far) but that the reader occasionally said the word "um" and did it in strings, sometimes 20 or 30 "um"s.
This string then cut into the same 5-7 word fragments by the same preschoolers and dumped into computer would not be solvable. The non-um part would be solved, and then there would be brackets for any um elements in excess of 7 words or so and it would simply be called a "um element" . There is no way with 5-7 word fragments to assemble conclusively a 20 um sequence. 20 ums and 2000 ums look identical.
One could count the number of pages in the encyclopedia, know the font size, and speak towards the total number of ums, but they cannot define the specific size of a particular um element.
Such is the case with things like the human genome.