Much of MARC coding is interpreted by bytes in position. In terms of
MARC tags/indicators, there are never more than 5 bytes or characters to
interpret, the first 3 being the MARC tag and the last two being the
indicators. The only way, IMO, to reliably encode these is to remove
all medial spaces (spaces in attribute values are not generally a good
idea anyways). We currently use such a system at our institution to
distinguish between Biographical (5450_) vs. Historical note (5451_).
So I can easily peek at what the first indicator is by checking the
fourth character in this string (in XSLT one could do this using the
substring() function). This helps especially in an EAD to MARC
conversion. I also don't see a problem, when both indicators are blank,
just encoding it like this: 845$a
If one is consistent in this approach, the fourth character would either
indicate the first indicator value or would return a subfield delimiter
>>> So, how does one represent
>>> 110 2_ $aMinnesota Historical Society
>>> in an encodinganalog in <origination><corpname> where the underscore
>>> represents a blank space?
> Maybe I'm missing something obvious but both these options seem to parse
> just fine.
> encodinganalog="110 2_ $a"
> encodinganalog="110 2 $a"
> What's the problem with the underscore or the space?