This is how I would (begin to) debug it.
 
REVIEW the list of subscribers.
Integrate that output in a listserv DISTRIBUTE MAIL DEBUG=YES
job.  Send the job to the LISTSERV hosting the list.
 
Review output.  Determine all LISTSERV sites responsible for
"exploding" a distribution.  Follow those LISTSERV sites via
RSCS/NJE Q commands and/or SMTP gateways (some sites have
multiple) making sure that any LISTSERVs along the path also
respond e.g., issue the THANKS command to them.
 
Observe any MAILER involvement (at the bitnet/internet gateway,
if applicable, or final LISTSERV "explosion" point), then SHOW
ROUTE (LMAIL) or SHOW DOMAIN (Crosswell).  Follow those
nodes/hosts.  When and if you actually come to Bitnet/Internet
gateway host, try to do a DiG of NSLOOKUP on all hosts and
authoritative names servers making sure there is consistency.  If
so, make sure that the "MX" (if any) or otherwise "A" host SMTP
ports respond appropriately and timely.
 
You may find that some IP address has changed too, yet the SMTP
client still has the old one cached.
 
                             -=-=-=-=-=-=-=-
 
In particular, if you find a bitnet distribution point of UKACRL
(aka ib.rl.ac.uk) you'll probably need to talk to the JANET
administrator US@UKACRL for assistance.
 
This is pretty much the general idea of what I learned while
running lists and administrating e-mail systems.
 
--  co-owner INFOSYS, TQM-L, CPARK-L, ERAPPA-L, JANITORS, LDBASE-L, et -L
[log in to unmask]        "I get paid by the Byte"        +1 814 863 1843
31 Shields Bldg. --   Penn State    -- University Park, PA 16802-1202 USA