In Judges 3:16, Ehud's sword or dagger was said to be a gomed in length. How long is a gomed? I'm aware some translations have cubit, but if it was a cubit, why wasn't the Hebrew word for cubit used?
The Hebrew word 'gomed' is used, in this verse, to apply to a double-edged dagger which is about a foot long. The NLT suggests it is about 12 inches long, whereas the NIV suggests 18 inches. The point is (sorry, no pun intended) this dagger could be strapped to a man's thigh. A dagger between 12 and 18 inches in length might work. Certainly, nothing longer.
Source: New Living Translation Study Bible notes (page 424).
The exact length of the sword is not significant for understanding the text. Translations may use whatever terms readers are expected to be familiar with. Since cubit is fairly well known term, many measures that aren't really cubits are probably translated as "cubit". Some translations even use feet and inches.
Cubit is an English word that refers to various forearm-based measures. What we call a cubit had different names and different lengths in different cultures. The lengths may be based on distances between different body parts, such as elbow-to-wrist, elbow-to-knuckles, or elbow-to-fingertips. Even measures with the same name in the same culture in different time periods had different lengths. Modern measures, such as the meter, have also been redefined several times.
There were two cubits: the standard cubit of 17.5 inches and a short cubit of 15 inches.
The גֹּ֫מֶד is a short cubit of 15 in. (about 38 cm).
Brown-Driver-Briggs גֹּ֫מֶד noun masculine cubit (from elbow to knuckles of clenched (contracted) hand; Aq on Ezekiel 27:11, Ra and others; Greek πυγμή; Late Hebrew גּוֺמֶד cubit; so Aramaic גַּרְמִידָא ChWB1. 155, (lexicons)) — Judges 3:16 of Ehud's sword אָרְכָּהּ ׳ג (see GFM).
The word in issue is גמד [GMD], instead, the commoner term for ‘cubit’ is אמה [AME]. According the traditional definition, but also on the basis of some archaeological confirms, the cubit was long 44,4 centimeters.
Now, for the common paradigm, the ‘aleph’ and the ‘he’ can be mater lectionis. So, they work sometimes as consonants and sometimes as vowels. If in this case they work as vowels, they do not carry any (absolute/unique) meaning, like that carried by the consonants, but only some cross-meanings (moods, diathesis, etc.), that is, a some meanings commons to every conceptual roots.
Consequently, the only letter (in אמה [AME]) that is re-echoing the original conceptual root is M (‘mem’), here working as medial radical. Then, there is the possibility that גמד [GMD] (also here the ‘mem’ works as medial radical) is the archaic term for ‘cubit’, derived from an omographic conceptual root.
This supposition is strenghtened by the Akkadian term for ‘cubit’, that is, AMMATU (‘forearm, cubit’, in ancient Assyrian/Babylonian) [Chicago Assyrian Dictionary, I:2:70-75]. Take account – please - the common (historical-ascertained) commutation between D > T.
Granted, some hypothize, instead, that GMD could be a different measure (a diminished cubit?), but their backing argument is not solid, because the manners to ideate alternative measurements are virtually a lot. Moreover, if GMD was a ‘short cubit’ “where else is this cubit mentioned?”, asks John Parkhurst (on גמד entry of his Hebrew and English Lexicon). In other words, if GMD was a ‘short cubit’, why we find – in MT - over 150 times the ‘cubit’ term, but only once the ‘short-cubit’ term? And where are the other various possible permutated measures among palms, spans, and cubits? If all those existed, where are they mentioned in MT?
However, there is a possibility that this term is present in MT twice, instead of only one occurrence (anyway, this fact does not modify the unbalanced ratio above mentioned between [standard]-cubit term and diminished-cubit term). In fact, in Ezekiel 27:11 we find the term גמדים [GMDIM, a simple plural of GMD]. Leaving out the ‘humorous’ Vulgate rendering ‘pigmei’ (do you picture a lot of ‘ferocious’ pygmies on the city towers?), we could to focus ourselves on an English term: ‘dagger’. Probably, the vulgar Latin term daca was the basis, passing through the Middle Low German term dagge, and perhaps also with a cross-derivation from the Old French term dague, resulting in the Middle English term dag, ‘to pierce’ (see also Middle Dutch dagge, Danish daggert, German degen, Old Norse daggardr, Walsh dager, dagr, etc.). So, they were not ‘pygmies’ but ‘dagger carriers’, or ‘daggermen’. I must specify what link may exist between GMD and 'dagger, since the two common consonants (G, D) are not written in the same order of pronounciation (G-D vs D-G). We have to remember that in the MT conceptual roots' corpus there are, in a lot of instances, roots with identical meaning although with a different permutation of the radicals (called 'allographic roots', technically) [some examples: 'To elevate' = MRA (Job 39:18) Vs RAM (Zec 14:10); 'To be stupid' = KSL (Jer 10:8) Vs SKL (Gen 31:28); 'To mourn, to weep tears' = BKA/BKE (Gen 21:16) Vs KAB (Job 14:22)]. Knowing this is easy to see the link above mentioned.
Moreover, we would consider also what we may draw from linguistic comparison. In Italian language, for an example, the word for ‘elbow’ is gomito. Would be only a coincidence the striking similarity between gomito and GMD (also in this case you have take to account – please - the common historical-ascertained commutation between D > T), along with the striking semantic affinity between ‘elbow’ and ‘cubit’?
In conclusion, it is more probable that גמד [GMD] indicated a cubit, in a more archaic term compared to אמה [AME]. On Judges 3:16 Keil & Delitzsch commented (bold is mine): “[Hapax legomena][גמד] signified primarily a staff, here a cubit, according to the Syriac and Arabic; not ‘a span’, σπιθαμή, lxx)”.