It isn't at all clear from your description what you are actually
hoping to use as your data source,  what the field names might be, and how
you are hoping to get the data into the resulting document. So I may have
misunderstood your question completely, but I hope the following makes
sense:

In essence, your data source needs to be a 2-D table (i.e. with columns and
rows, with a field name for each column and, usually, the same number of
columns in each row). That might be implemented in a number of different
ways, e.g.
a. a comma-delimited text file with e.g.

fieldname1, fieldname2, fieldname3
row1field1value, row1field2value,row1field3value
row2field1value, row2field2value,row2field3value
row3field1value, row3field2value,row1field3value
etc.

b. the same thing in a Word document
c. a table in a Word document
d. an Excel worksheet
e. a table or query in an Access database
f. a table or view in any database you can access via ODBC
g. separate data and header fiels are allowed for some types of file
..etc.

What it cannot be is a disconnected recordset or some structure purely of
your own invention.

What you need to get Word mailmerge to merge the data in from that data
source is to use { MERGEFIELD } fields whose names correspond to the columns
in your table, e.g.

{ MERGEFIELD fieldname1 }

You would either have these mergefields in your Word document already, or
use the methods of WOrd's MailMerge object to insert them.

For everything else, you need to have a look at Word's MailMerge object's
methods, properties etc.

Alternatively you can avoid Word's built-in MailMerge altogether and "roll
your own" merge, but then it's up to you how to decide where to put each
field in your data source, etc. etc.