I don't believe that the above analysis is correct. Please look at the
Gemara (Tslmud) in Avoda Zara 9a and ff. Note how even Rashi printed the
alef in front to designate 1000, the dalet to designate 4000. From what I
know from my own teachers we have always designated the thousands. If for
some reason this is not clear in some places, I think that it is due to the
same fact that most people call this 97 not necessarily 1997 - the hundreds
and thousands are obvious and known to all. In fact it is so commonplace
throughout the world to drop-off the thousands that the entire "year 2000
problem" would not be around if it weren't for this fact. Sometimes in
books you may find that the publishing year has been turned into an
interesting Gematria (using the numerical value of the letters to deepen
understanding of the word or concept) which has to do with the title in
some form. In such cases the Gematria is usually for the year ( i.e the
Hebrew letters representing the year) w/o the thousands, and therefore is
proceeded with a peh " kuf which means lepkidah ketana ( a small counting
i.e w/o the thousands).
As for the Ketuva (marriage contract) I don't believe that the translation,
though technically correct, reflects the meaning of the words in Hebrew: in
Hebrew it is: leminyan she-anu monim kan be... I think a more accurate
translation is simply: 'according to the count [of the years]that we count
here. The reason for stressing the "here" is because Jewish communities
used to use a number of different methods to count years the most famous of
which is minyan "shtarot" (contracts) which started in the time of the
Ptolemaic kings in Egypt in the 1st century BCE, and continued to be used
until the time of the Rishonim ( early Talmudic commentators 1000 - 1500 CE
approx). Since a ketuva is first and foremost a "shtar chov" - a debit note
- the date is essential and must be exact.
Hope this clarifies some of the points mentioned.
Moshe Genuth.