<?tei:SOF p2lm.odd ?>
<!-- Revisions:                                               -->
<!-- 93-06-02 : MSM : change N value                          -->
<!-- 93-06-02 : MSM : new intro para post Eynsham             -->
<!--                  revert to old time element              -->
<!--                  merge all name-clashes with core        -->
<!-- 93-06-02 : MSM : delete time-expression sectn per DGR    -->
<!-- 93-04-12 : MSM : change time to dgrtime                  -->
<!--                  clear name clashes with address,        -->
<!--                  region, postbox, dateline, country      -->
<!-- 9 apr: changed ext to extn; title to title.pr to avoid confusion -->
<!--        added dtdfrags   -->
<!-- ODD checked: LB: 29 Mar 93 -->
<!-- From: Dan Greenstein <digger@uk.ac.glasgow.udcf>   -->
<!-- Date: Tue, 7 Jul 92 14:16:24 BST   -->
 
<div1 id=LM n=12><head>Letters and Memoranda</head>
 
<p>This section attempts to identify the characteristics of two
document types produced  and exchanged in an ordinary office
environment.  Its recommendations are relevant both to those
interested in implementing these Guidelines in a working office
environment and to those who will use them to encode such
documents for study, manipulation or analysis.  Of course,
letters and memoranda are not static textual objects.  Their
content changes with some frequency as they circulate a working
office environment.  Where such changes have been implemented,
they appear in the historical record as manuscript editions,
marginal comments, glosses, etc.  Consequently, with some
applications it will be necessary to pay close attention
recommended in these Guidelines for recording such glosses and
alterations.  In particular, the reader is recommended to consult
the sections on simple editorial interventions (<xref target=COED>),
notes and annotation (<xref target=CONO>), and transcription
problems (<xref target=PH>).
<!-- Michael, I am not au fait with your xref targeting system-->
<!-- so have provided the numeric headings as they appear in P2-->
<!-- 6.3.5 (simple editorial changes) 6.8 (Notes, Annotation..) -->
<!-- Section 21, Transcription etc -->
 
Where it is appropriate to mark referencing strings (e.g. references,
not necessarily by means of proper nouns) to objects,
people, places, or to mark proper names of places, persons and
organizations, and dates, see section <xref target=CONA>
and chapter <xref target=ND>.
 
<!-- your reference to names dates and places -->
 
The new elements expressly used in letters and memoranda and
introduced in this section include:
 
 
<tagList>
<tagDesc tagDoc=SENDER>
<tagDesc tagDoc=RECIP>
<tagDesc tagDoc=SIGNAT>
<tagDesc tagDoc=PREPARER>
<tagDesc tagDoc=SENSIT>
<tagDesc tagDoc=STATUS>
<tagDesc tagDoc=PSCRIPT>
<tagDesc tagDoc=DATELINE>
</tagList>
 
<p>Also proposed in this section are a number of <term>crystals
</term> suitable for use in office documents linked with database
management and information retrieval systems, in which detailed markup
of proper names (organizational, geographical as well as personal
names), addresses (postal and telecommunications), meetings and events
can be useful. These crystals include:
 
<tagList>
<tagDesc tagDoc=ADDRESS>
<tagDesc tagDoc=TELECOMM>
<tagDesc tagDoc=EMAIL>
<tagDesc tagDoc=ACTION>
<tagDesc tagDoc=DOCREF>
</tagList>
 
and mention is made as well of crystals described elsewhere including:
 
<tagList>
<tagDesc tagDoc=PERSNAME>
<tagDesc tagDoc=PLACNAME>
<tagDesc tagDoc=ORGNAME>
<tagDesc tagDoc=DATE>
<tagDesc tagDoc=TIME>
</tagList>
 
 
 
<p>From a general point of view an office document consists of three
parts: front matter, body matter and back matter. The text body, apart
from the subject and/or dateline, salutation and signoff features is
similar to the body of other document types making repetition of the
relevant core and base tags discussed in chapters <xref target=CO> and
<xref target=DS> respectively redundant here. More interesting and
characteristic of this document type is its front and back matter which
can be seen as a sort of document profile. Both contain information
 related to various special aspects of the document including its:
 
<list>
 
<item>preparer(s), author(s), sender(s) and intended
recipient(s) who will be identified often alongside their respective
postal, telecomm and/or e-mail addresses
 
<item>storage (working office systems will record local file
names, access authorizations, etc)
 
<item>status (e.g. draft, private and confidential, internal) and
revision history
 
<item>distribution by post or electronic mail (identifcation of
originator, authorizing person, primary and secondary recipients, etc)
</list>
 
Most of this section is therefore devoted to the front- and back-
matter and takes into account ISO standards (ODA document profile) as
well as CCITT norms (X400 series). It also builds upon the results of a
number of projects of the Commission of the European Communities in the
field of document filing and archiving.
 
<div2 id=LMFRONT>
<head>Front and back matter of letters and memoranda</head>
 
<p>In either the front or back matter of a letter or memorandum is
contained the names (individual or institutional) and addresses (postal
and telecommunication) of the person(s) and/or institution(s) who sent,
prepared, authored, and signed the document and the name(s) and
address(es) of the document's intended recipient(s). To identify the
contributing sources and recipients of such a document, the following
elements are recommended:
 
<tagList>
<tagDesc tagDoc=SENDER>
<tagDesc tagDoc=RECIP>
<tagDesc tagDoc=SIGNAT>
<tagDesc tagDoc=PREPARER>
</tagList>
 
<p>The sender is conceived of as the name and/or address of the document's
ultimate source of authority.  The preparer comprises the name and/or
address of whoever (or whatever group) was responsibility for typing or
writing the letter or memoranda perhaps on behalf of the author. The
signatory comprises the name and/or address of the letter's or
memorandum's signer and might be different from both the author and the
preparer. And the <gi>recipient</gi> comprises the name and/or address
of the intended recipient(s) of the letter or memorandum. Each of these
elements may contain names (personal or organizational) discussed in
<xref target=ND> and/or addresses (postal, telecommunication
and email) discussed below.
 
<p>In some cases, names and addresses of senders, recipients, etc. are
separated physically from one another in a document making their
alignment an interpretive gesture. Where this is the case, one of the
aligning mechanisms described in section <xref target=SA> is
recommended. Alternatively, distinctive parts of the <gi>sender</gi>
<gi>recipient</gi> etc., elements may be identified through the numeric
values of their <att>part</att> attributes.
 
<p>The <gi>recipient</gi> and <gi>sender</gi> elements also share a
<att>type</att> attribute with two recommended values &mdash;
primary and secondary &mdash to distinguish between primary and secondary
recipients and senders, respectively, where more than one is present.
Secondary authors include functionaries asked to write a letter and have
it sent on behalf of some governing individual or body. Secondary
recipients include the named individuals or departments to whom letters
and memoranda are sent care of. A letter sent to John Doe care of Mr and
Mrs Anthony Doe, Mr and Mrs Anthony Doe are both secondary recipients.
Where letters or memoranda are jointly authored or sent to several
recipients each author or recipient should be treated as primary perhaps
in a <gi>list</gi> element (see section <xref target=COLI>).
 
<p>Where a letter is concerned it may be important to distinguish between
that information which is provided for the <gi>sender</gi> or
<gi>preparer</gi> because it is printed on the stationary on which the
letter is sent and that which is provided in the author's or preparer's
hand (or typewriter/printer). Stationary <q>borrowed</q> from a
Holiday Inn but used for personal correspondences
can, for example, provide misleading. Printed and entered information
can be distinguished by using printed and entered as recommended values
of the <att>imprint</att> attribute of the <gi>sender</gi> and
<gi>preparer</gi> elements.
 
<p>Other features commonly found in both the front and back matter of
letters and memoranda include indications of their sensitivity (e.g.
<q>private and confidential</q>) and their draft status
(e.g. <q>draft version 2</q>). Here the <gi>sensit</gi> and
<gi>status</gi> elements are recommended.
 
<p>References to the current document and/or to other documents (including
letters, memoranda, reports, etc) are also frequently occurring features
in the front- and back-matter of letters and memoranda but also in their
text body. Few will be unfamiliar with the text strings <q>Our Ref:</q>
and <q>Your Ref:</q> which appear regularly in the headers of of
official correspondences. Invariably, the references filled in for
<q>Our Ref:</q> are incomprhensible, but may prove important where a
corpus of correspondence is being compiled for a critical edition, an
archival catalogue, or for historical investigation for example. To tag
such references, wherever they appear, the <gi>docRef</gi> element is
recommended. Its <att>whose</att> attribute is used to indicate the
owner of the reference scheme and may be used to point at a sender,
recipient or any other person or organization identifier established via
an appropriate alignment mechanism (see section <xref target=SA>).
 
<p>The <att>system</att> attribute may be used where the referencing system is
identifiable and might include values such as <q>Canadian Civil
Service</q>. The <att>what</att> attribute indicates what precisely is
being cross referenced. Values here may include other document types
(e.g. letter, memorandum), accounts (e.g. bank accounts, credit card
accounts). Where cross-references are found in the text body, they are
just as likely to be to events (<q>With regards to your phone call the
other day</q>; <q>it was good to bump into you on the street</q>). Values
indicating events or types of events might prove invaluable to the
diplomatic or political historian who will want to know, for example,
what led Nixon to do x and will search his letters, etc for some
reference to possible stimuli.
 
<!-- LOU AND MIHAEL - NB IT IS ESSENTIAL to come to some conclusion about   -->
<!-- whether something like a citation or docref can be made to events as   -->
<!-- well as actual documents. My suspcion is that the mechanisms already   -->
<!-- exist in (a) cit tag and (b) the material used in the header of aural   -->
<!-- and visual media or at least (c) span, but if (a) or (b) then these will   -->
<!-- need some extension. This is not just being picky. We need some   -->
<!-- mechanism to mark references which do not refer to texts or some other   -->
<!-- natural language recording(audio, etc).  I hesitate to develop yet   -->
<!-- another mechanism here but have done so hoping you'll chuck it out.   -->
 
<p>Anything in the front-matter may also appear in the back- matter of a
letter with one addition, that is, postscripts which may be tagged with
the <gi>postscript</gi> element.
 
<dtdFrag id=DLMFRONT n='Elements for front matter of Letters and Memos'>
<tagDecl tagDoc=SENDER>
<tagDecl tagDoc=RECIP>
<tagDecl tagDoc=SIGNAT>
<tagDecl tagDoc=PREPARER>
</dtdFrag>
 
<div2 id=LMBODY><head>The Text Body of Letters and Memoranda</head>
 
<p>Letters and memoranda are frequently only a part of an ongoing
discourse. Consequently they may take on a tripartite structure which
it may be desirable to tag using the <gi>s</gi> or <gi>span</gi>
elements discussed in sections <xref type='div1' target=SA>.  In a
letter or memorandum an introductory section opens with salutations and
often a date or dateline before developing with cross references to
other documents and/or events which form the history of the particular
discourse which has led to the current document. These references might
be tagged with the <gi>docref</gi> element discussed above. A
substantive section then follows which grows out of the documents and
or events cross-referred in the introduction. A conclusory section
points to future action on the part of the letter's author(s) and or
recipient(s) (perhaps with other cross-references) before closing with
final salutations and signatures.
 
<p>A letter from Edmund Burke to Lord Grenville dated September 19, 1792
may be used as example. It opens with salutations and a dateline
comprising the date and place of the letter's composition. The text then
begins with references to a previous letter to Grenville containing
Burke's thoughts on the "probable effect of the French Revolution on the
whole of Europe", and to a letter containing Grenville's response
thereto. The body then develops Burke's current thinking on the issue
and the conclusion comprises a request that the discussion be taken up
further preferably through a personal meeting. The letter is then
formally closed with additional salutations and a signature. A formal
bill or demand for payment is no more than another particular instance
of this same general structure. The introductory section comprises
references to charges made against the account of the bill's recipient.
The body includes the substantive issue of present concern (normally a
brief summary of how much is owed) and a conclusion which asks for
payment and explains how it might be made and what might happen if it is
not.
 
<p>Where it is desirable to indicate some of these features the
following elements are recommended:
 
<tagList>
<tagDesc tagDoc=SALUTE>
<tagDesc tagDoc=HEAD>
<tagDesc tagDoc=DATELINE>
</tagList>
 
The <gi>salute</gi> element is described in section <xref target=DS>.
In a letter it may be used to tag opening and closing greetings, for
example:
 
<eg><![ CDATA[
My dear Florence the first,
Empress of Scavengers
Queen of Nurses
Maiflextresse of droflexlesses
Governess of the Governors of India
Reverend Mother Superior
Mother of the British Army
&c. &c. &c.
]]>
</eg>
 
which begins a letter of 9 November 1868 from Benjamin Jowett to
Florence Nightingale
 
<eg><![ CDATA[
or I have the honour to be, with real regard and esteem, Dear sir, Your
most faithful and obedient humble servant,
]]>
</eg>
 
which appears at the end of a letter from the Rt Hon,. Edmund Burke to
Joseph Bullock, Esq of 3 March 1783
 
<p>In letters and memoranda, the date and/or place of composition (as
distinct from the sender's address) is often indicated, sometimes in the
front matter but with letters, normally on the same line as the opening
salutation and as such is considered part of the text body. Examples
include <q>October 1868</q> and <q>Launceston, Monday
morning, March, 1783</q>. These may be tagged simply with the
<gi>dateline</gi> element which may be used to enclose
tagged <gi>placname</gi>, <gi>date</gi> and <gi>time</gi> elements as
in:
 
<eg><![ CDATA[
<dateline>
     <settle type=village>Launceston</settle>,
     <dayname>Monday</dayname>
     <timeName>morning<timeName>,
     <month>March</month>,
     <year>1783</year>
</dateline>
]]>
</eg>
 
<p>The <gi>head</gi> element has already been discussed in section
<xref  target=DS> and may be used to tag headings
which appear on the lines immediately below or above the salutation
and/or dateline and provide a brief synoptic statement concerning the
document's contents e.g. <q>Personal Current Account No
00162451</q>.
 
<p>Signatures normally appear after closing salutations and may be
tagged with the <gi>signature</gi> element introduced in section <xref
 target=DS>. To distinguish handwritten signatures from their
more formally typed presentation (both frequently occur in letters),
printed and entered are the two recommended values of element's
<att>imprint</att> attribute. Where desirable, the component parts of
the personal names represented in the signature may be tagged with
<gi>persName</gi> and its associated elements. In addition, the
<att>whose</att> attribute enables the encoder to point at a sender,
recipient or preparer ID.
 
<p>It may also be important to distinguish between printed information and
that entered by the letter's author or preparer with printed form
letters which are tailored by the author or preparer by deleting all but
the most appropriate options (e.g. <q>Dear Sir/Madam</q>) or by filling
in spaces left blank on the form. In either case, printed information
may be distinguished from entered information using the <gi>span</gi>
element discussed in section <xref  target=AI>.
 
 
<div2 id=LMCRYS><head>Crystals Common to Letters and Memoranda</head>
 
<p>Many of the crystals already discussed elsewhere in this volume will be
found in letters and memoranda including:
 
<tagList>
<tagDesc tagDoc=PERSNAME>
<tagDesc tagDoc=ORGNAME>
<tagDesc tagDoc=PLACNAME>
<tagDesc tagDoc=DATE>
<tagDesc tagDoc=TIME>
<tagDesc tagDoc=MEASURE>
</tagList>
 
Others are introduced in the immediately preceding sections
including:
 
<tagList>
<tagDesc tagDoc=DOCREF>
<tagDesc tagDoc=DATELINE>
</tagList>
 
In addition, the following crystals are introduced here:
 
<tagList>
<tagDesc tagDoc=ADDRESS>
<tagDesc tagDoc=TELECOMM>
<tagDesc tagDoc=EMAIL>
<tagDesc tagDoc=ACTION>
</tagList>
 
<div3 id=LMCRYS1><head>Postal Addresses</head>
 
<p>Postal addresses are features which appear in many document types and
are especially important in letters and memoranda. They should be tagged
where desirable with the <gi>address</gi> element. The following
elements are also recommended for tagging the addresses component parts:
 
<p>Addresses may appear as features of the <gi>recipient</gi> and
<gi>sender</gi> elements, for example, or independently
within the front or backmatter of letters and memoranda. They may be
tagged simply with the <gi>address</gi> element. Where it is desirable
to indicate the component parts of an address the following tags are
recommended:
 
<tagList>
<tagDesc tagDoc=ADDRESS>
<tagDesc tagDoc=PERSNAME>
<tagDesc tagDoc=ORGNAME>
<tagDesc tagDoc=LOCNAME>
<tagDesc tagDoc=APTPOS>
<tagDesc tagDoc=STNUM>
<tagDesc tagDoc=STNAME>
<tagDesc tagDoc=POSTBOX>
<tagDesc tagDoc=PLACNAME>
<tagDesc tagDoc=POSTCODE>
</tagList>
 
<p>Organizational and personal names should normally be treated as separate
crystals rather than as part of an address which is, strictly speaking,
a locational device. There, are however, three exceptions:
 
<list>
 
<item>An organizational name, for example, may be considered
as part of an address where the address pertains to a person (whose name
should normally be tagged as a <gi>persName</gi>) and merely helps
locate the person in space (and presumably as well for the local post
office) e.g. <q>Sir Keith Thomas, The British Academy</q>.
<item>A personal name may appear as part of an address where
it is the name of a preson acting for someone else as a recepticle for
items sent to the address. In an address which begins
<q>John Doe, care of A.N. Other</q>, A.N. Other might be included as a person's
name as part of the address element.
<item>Where addresses appear in the front- or back-matter of
letters and memoranda alongside the names (personal or organizational)
of their senders or recipients, both addresses and names should be
tagged as elements which are part of the <gi>sender</gi> or
<gi>recip</gi> elements, respectively as discussed elsewhere in this
section. </list>
 
<p>Where it is desirable to tag the constituent parts of an address, the
following tags are recommended. The <gi>locName</gi> comprises a name or
proper name of the property whose postal address is indicated e.g.
<q>Tara</q>. The <gi>aptPos</gi> element, on the other hand,
indicates the position of the particular apartment, normally in relation
to apartments normally within the same building or complex e.g.
<q>top left</q> <q>second right</q>
or simply <q>#3</q>.  The <gi>streetName</gi> and
<gi>streetName</gi> elements are recommended where it is
desirable to tag street names and street numbers.  The <gi>postBox</gi>
element is self explanatory and need only include the postbox reference
tag or number.  Geographical place names are notoriously difficult to
standardize where they occur in addresses as elsewhere given the
variation in the geopolitical sub-divisions used therein.  Consequently,
in addition to the relatively simple elements <gi>city</gi> and
<gi>region</gi>, and <gi>country</gi>,
two somewhat more general
elements (<gi>settlement</gi> and
<gi>bloc</gi>) have been provided for this purpose. They may be tailored
more closely to the address at hand with the string values entered in
their <att>type</att> attributes. Lastly, the <gi>postcode</gi> element
is used to tag postcodes where this is desirable. Numeric values entered
into its <att>part</att> attribute facilitate tagging multi-part
postcodes, such as those split by city names in German addresses, for
example.
 
 
<div3 id=LMCRYS2><head>Telephone and Fax Numbers</head>
 
<p>These appear regularly in the front and/or back-matter of letters and
memoranda and may appear as well in the text body. The
<gi>telecomm</gi> element is recommended where it is
desirable to encode phone and fax numbers which are distinguishable by
the character value entered into the element's <att>type</att>
attribute. For database-type applications where it is useful to encode
the component parts of the telephone or fax number, the following
elements are recommended:
 
<tagList>
<tagDesc tagDoc=CCODE>
<tagDesc tagDoc=AREACODE>
<tagDesc tagDoc=TELNUMB>
<tagDesc tagDoc=EXTN>
<tagDesc tagDoc=INTERNAL>
</tagList>
 
 
<p>The <gi>countryCode</gi> element is used for the country code (only
applicable where numbers are given for use by callers from abroad). The
<gi>areaCode</gi> element is used for regional or area codes. The
<gi>telNumb</gi> element encloses the local telephone or fax numbers.
Where these appear in two parts separated by a hyphen, space or other
diacritic, the parts themselves may be significant (representing a
telephone exchange, for example, in the U.S. and parts of the U.K.) and
may be distinguished from one another by the numeric values in the
<att>part</att> attribute of the <gi>telnumb</gi> element.  The
<gi>extension</gi> element is used for extension numbers and the
<gi>internal</gi> element is used where institutional telephone systems
are in place providing shorthand or <q>internal</q> numbers which are
from or subsections of the <q>external</q> number.
 
<div3 id=LMCRYS3><head>E-mail Addresses</head>
<!-- It was recommended that the section on e-mail addresses -->
<!-- also be changed. Please use the following introduction if -->
<!-- possible -->
 
<p>E-mail addresses are used to route a text message in electronic
form from a sender on a network to one or more recipients on the
same or different networks.  When a message has to travel across
network boundaries, the address's structure is used to
facilitate the appropriate navigational path.  Such addresses are
not standardized and tend to differ from network to network.
<!-- Consequently, we recommend that                          -->
 
<!-- Steve Derose recommended that e-mail addresses be        -->
<!-- declared as a notation and declared in the header. Can   -->
<!-- you supply the appropriate documentation?                -->
<!-- not now. -msm                                            -->
 
Consequently, the following elements represent only a tentative set
of recommendations, which will need to change to reflect changes in
the network infrastructure.
 
<tagList>
<tagDesc tagDoc=EMAIL>
<tagDesc tagDoc=NETNAME>
<tagDesc tagDoc=NODENAME>
<tagDesc tagDoc=USERNAME>
<tagDesc tagDoc=ROUTING>
</tagList>
 
<p>The <gi>email</gi> element is used to enclose the entire address. The
remaining recommended elements are used to encode component parts as
follows:
 
<list type=gloss>
 
<label>netname<item>network name or addressing style (e.g.
<q>EARN</q>, <q>JANET</q> etc.)
<label>nodename<item>(e.g. <q>UK.AD.GLASGOW.UDCF</q>,
<q>flash.bellcore.com</q>, etc.)
<label>username<item>user name, ID, account (the individual
identifer within the node whose format is locally determined.
<label>routing<item>(for cases where the sender must specify
a routing for mail </list>
 
Examples include:
 
<eg><![ CDATA[
The Text Encoding Iniative discussion server is located at:
<email>
     <username>TEI-L</username>
     <nodename>UICVM.CC.UIC.EDU</nodename>
     <netname>Bitnet</netname>
</email>
 
or
 
<email>
     <username>TEI-L</username>
     <nodename>UICVM</nodename>
     <netname>Internet</netname>
</email>
 
Subscription requests reading <q>subscribe TEI-L [name of
subscriber]</q. should be sent to
 
<email>
     <username>Listserv</username>
     <nodename>UICVM</nodename>
     <netname>Internet</netname>
</email>
]]>
 
</eg>
 
<div3 id=LMCRYS4><head>Action Points</head>
 
<p>
Indications of actions taken, required or suggested may crop up within
the front-, body- and back matter of letters and memoranda. Here the
enormous amount of variation in the nature of the action required and by
whom make any more than a generic <gi>action</gi> element with an
open-ended <att>type</att> attribute entirely impracticable. Where
processing applications require that rather more information about the
nature of actions be recorded (e.g. the results of an open or roll-call
vote taken of members present at a particular meeting), it is advisable
to use feature structures to serve this purpose (see section
<xref  target=FS>).
 
<!-- EDS - Here is a brief section for time. It should -->
<!-- probably be placed in section 23 next to names, dates -->
<!-- and measures but I shall of course leave this up to you.-->
 
<?tei:EOF p2lm.odd ?>
