My experience of having read tens of thousands of historical documents 
in court and secretary hand tells me that the chances of finding a new 
word or artefact are virtually zero. These difficult words are 
inavariably spelling variations or as Tasha points out 
mistranscriptions. It is thus far more likely that  a pair of shillards 
is a pair of stilyards than an unknown artefact. Obviously you can't 
prove it but this isn't physics - as I say try transcribing a medieval 
latin document where the words are abbreviated and m, n, u and i are all 
reduced to minims (iii etc), you have to rely heavily on experience and 
knowing what a word is likely to be. Context within a documenta and 
comparison between documents is the usual key method. Caution is a good 
thing but if you rely on proof we might as well throw most historical 
documents prior to printing in the fire. If I was using this in an 
article I would quote the original form as a stronger element of 
interpretation than normal is present but it may well be a transcription 
problem anyway. Even skilled paleographers have to make intelligent 
guesses in separating many letter forms, and in medieval latin one of 
the first things you are taught are likely variant spellings otherwise 
using a dictionary is impossible.

Ron May wrote:

>So what you are saying, Paul, is that somewhere you have seen a definition of 
>"shillard" that matches "Steelyard"? And/or other scholars have analyzed the 
>term and come to the same conclusion? I am sure most folks would find that 
>strong evidence. However, I recall an instance in which someone here in 
>California mis-translated "casamata" at the Royal Presidio de San Diego to mean 
>"butcher shop/butchering house," only to find that the association with the Spanish 
>Army fort meant, in this case, "casemate" or powder and ball storage building. 
>It is for this reason that I am cautious with jumping to conclusions.
>
>Ron May
>Legacy 106, Inc.
>
>  
>