Halakhah can be minutely precise. So it seems, and so it is. Thus it is interesting when it fails to be precise, especially in an area where one would expect rigorous precision. Particularly in an area where not only does one expect rigorous precision, but rigorous precision is actually mandated. I mean, specifically, the halakhot of writing a Sefer Torah, a Torah scroll.
Here is an excerpt from the Anchor Bible Dictionary Vol. IV (K-N), pg. 212, entry "Language (Hebrew)" by Gene M. Schramm:
"In copying a Torah scroll, the scribe was bound to copy his text as a virtual facsimile of all other authorized Torah scrolls, with leeway to do little more than reduce or enlarge the page size of the parchment leaves. Everything else had to be true and exact and scaled to the template. The copyist faithfully began each line with the same first word, ended each line with the exact last word, and included the same number of lines and the same blank spaces which indicated the pericopes. An oversized letter was reproduced as oversized, a miniscule was copied as a miniscule. Upside-down letters, backward letters, and flawed letters--all were faithfully reproduced as seen. Even the mysterious dots that appear here and there over letters were copied. A Torah scroll that failed to be a faithful replica of its authentic prototype was unfit for use and had to be duly corrected."
This is actually a good description (as an aside, Schramm uses a good analogy in describing masoretic notes. He compares them to the modern editorial comment sic). But as you surely know, that which I highlighted in red is a mistake.
Indeed, there are conventions with the force of law designed to produce highly faithful reproductions of other kosher Torah scrolls. The general thrust of this paragraph is true. But the halakhah actually does not mandate which word begins and ends lines, only that a line includes a defined range of letters, namely it must have enough room to write the longest word in the Torah, לְמִשְׁפּחֹתֵיהֶם (Gen 8:19) three times. This obviously leaves room to write a range of words on each line, and scribes may write more or less in the space of that size. (Exceptions, of course, are where a word ends a pericope called pesuha, for the next word must begin on a new line, and the poetic songs in Exodus and Deuteronomy, which require a specific formatting.)
In addition, no specific number of lines per column, or columns per scroll is mandated, something which I first realized after seeing a very old Torah on display at the Jewish Museum. Knowing that there are 42 lines per column, I was surprised to see that this one had many more. This led me to learn Torah and discover that, indeed, the number is not standardized by halakhah.
It is, however, standardized by convention (why do you think I thought it had to be 42 lines? It's because just about all Torahs use 42 lines per column!). From the Encylopedia Judaica entry Sefer Torah:
Although there is no law regulating the number of pages or columns a Torah must have, from the beginning of the 19th century a standard pattern of 248 columns of 42 lines each was established.
(The above was written by Louis Rabinowitz and R. Aharon Rakefet-Rothkoff, back when he was Rabbi Dr. Arnie Rothkoff, rabbi and Wissenschafty type. ;)
I assume (should I?) that this practice is in R. Shlomo Ganzfried's Keset ha-Sofer (1835), which has been enormously influential in standardizing those areas of safrut which remained unregulated.
And, of course, although there are detailed instructions for adorning the letters with תגין, they are not מעכב.