The first example is from Thomas Blount's Glossographia (1656). The textual divisions are the dictionary (<ttdv1>), the alphabetical letter heading (<ttdv2>), the entry (<ttdv3>), and the main lexical subdivision within an entry (<ttdv4>). This subdivision will vary from one dictionary to another and is subject to editorial interpretation, so that there is no one right way to tag it. In the example from Blount, the entry contains only one <ttdv4>, a lemma or dictionary headword. Within this, only two tags appear, <form> and <explan>. The <f>, <sic>, <cit>, <term>, and <lang> occur unpredictably.
Note also that <term> may enclose words cited as objects in their own right, <egt> an example quotation or sentence (cf. TEI P3 <eg>), <egtr> translation of an example <cf. TEI P3 <tr>, pp. 1202-03), <etym> etymological remarks, <gen> notes of the gender of words, <i> inflections, <pos> the part-of-speech, <usg> a note of usage, and <xref> a cross-reference.
TEI P3 employs the floating tags <entry> and <entryFree> (pp. 958-59) instead of resorting to divisional tags. TEI <def> (definition; pp. 921-22) and <sense> (pp. 1146-47) are anachronistic in this period, which is well before people understood fixed referential definitions. Other TEI tags, such as <gram> (pp. 1002-03), <hom> (pp. 1014-15), and <stress> (pp. 1169-70), reflect modern lexicographical practice, which employs relatively strict fields of information in all entries.
Despite the apparent nesting here, it would be wise to leave the <ttdv?> tags unnested in any DTD.
Note that it is important to give the normalized form of the dictionary headword as the lemma attribute in the entry tag <ttdv3>.
<ttdv1 type=""dictionary">
...
<bkdv4 type="line" n="1"><ttdv2
type="alpha">A</ttdv2></bkdv4>
<p>
<bkdv4 type="line" n="2"><ttdv3 type="entry"
lemma="a"><ttdv4 type="lemma" n="1"><form
n="1"><f type="bkbl">A<f type="i"></form>
<explan>Alpha<f type="r"> is a Greek</bkdv4>
<bkdv4 type="line" n="3">privative (which</bkdv4>
<bkdv4 type="line" n="4">being set before</bkdv4>
<bkdv4 type="line" n="5">any simple word</bkdv4>
<bkdv4 type="line" n="6">deprives it of its</bkdv4>
<bkdv4 type="line" n="7">proper signification) and <sic
corr="sig{\-}">sig</sic></bkdv4>
<bkdv4 type="line" n="8">nifies as much as, without. </bkdv4>
<p>
<bkdv4 type="line" n="9"><cit author="Cicero"
work="???"><f type="i">Cicero<f type="r">
calls <term>A</term> <f type="i"><lang type="l">
literam saluta</bkdv4>
<bkdv4 type="line" n="10">rem<lang type="e">
<f type="r">, a comfortable Letter, be{\-}</bkdv4>
<bkdv4 type="line" n="11">cause it was a note of <sic
corr="Abso{\-}">Abso</sic></bkdv4>
<bkdv4 type="line" n="12">lution, but <term>O</term>,
<f type="i"><lang type="l">literam tristem<lang
type="e">.</explan></bkdv4>
</cit>
</bkdv3>
<bkdv3 type="col" n="2>
<bkdv4 type="line" n="1">a sorrowful letter, because it</bkdv4>
<bkdv4 type="line" n="2">denoted condemnation. See</bkdv4>
<bkdv4 type="line" n="3"><f type="i"><lang
type="l">Ignoramus<lang type="e"><f type="r">.</bkdv4>
</ttdv4>
</ttdv3>
The second example is from Randle Cotgrave's French-English dictionary (1611). This example shows two kinds of <ttdv4> units within an entry, one for the main headword, another for a phrasal variant of it. Note also how the dictionary headword or phrase has a richer variety of subdivisions: the gender <gen>, the exemplary quotation <egt>, the translation of that exemplary quotation <egtr>, and a usage statement <usg>.
<ttdv3 type="entry" lemma="abandon">
<ttdv4 type="lemma"><form n="1">Abandon:</form>
<gen>m.</gen> <explan> bandon, free licence, full libertie for o{\-}
thers to vse a thing; The quitting, abandonment, or
prostitution of a thing vnto others; </explan> whence;
<egt n="1"> Mettre sa forest en abandon; </egt>
<egttr n="1"> to lay it open, or
make it common, for all men that will gather wood, or
graze their cattell, in it. </egttr>
</ttdv4>
</ttdv3>
.....
<ttdv3 type="entry" lemma="abatteur">
<ttdv4 type="lemma"><form
n="1">Abbateur.</form> </ttdv4>
<explan> a feller, ouerthrower, beater, breaker, or
bearer, downe of. </explan>
</ttdv4>
<p>
<ttdv4 type="phrase" n="1"><egt n="1"> Grand
abbateur de bois. </egt> <egttr n="1"> a sore fellow, </egttr>
<egttr n="2"> horrible
swaggerer, </egttr> <egttr n="3"> terrible bugbeare; </egttr>
<egttr n="4"> one that ouerthrowes
all he meets with; </egttr> <usg type="style"> (Ironically.) </usg>
</ttdv4>
</ttdv3>
6.11.2 COCOA
Renaissance dictionary entries are not as clearly structured as those in dictionaries since the 18th century. Synonyms, comments on grammatical function, usage, or meaning, and general information, combine in ways that are untidy in the eyes of modern lexicographers. For this reason RET tags are kept general. Each entry is introduced by a <ttdv2> tag and then a series of <tt> tags for the form (lemma, or dictionary word-form), gender <tt gen>, inflection <tt i>, explanation <tt explan>, exemplary quotation or text <tt egt>, translation of the exemplary text <tt egttr>, and usage <tt usg>. Sub-entries, which often concern phrases built from lemmas, have the <ttdv3 phrase> tag. No attempt is made to make distinctions among different senses in the explanatory part of the entry. It is not clear that the Renaissance understood that words might have in themselves different senses.
<ttdv2 entry>
<lemma abandon>
<tt form> Abandon: <tt gen> m. <tt explan> bandon, free licence, full libertie for o
thers to vse a thing; The quitting, abandonment, or
prostitution of a thing vnto others; {whence;}
<tt egt> Mettre sa forest en abandon; <tt egttr> to lay it open, or
make it common, for all men that will gather wood, or
graze their cattell, in it.
.....
<ttdv2 entry>
<lemma abatteur>
<tt form> Abbateur. <tt explan> a feller, ouerthrower, beater, breaker, or
bearer, downe of.
<ttdv3 phrase>
<tt egt> Grand abbateur de bois. <tt egttr> a sore fellow, <tt egttr>
horrible
swaggerer, <tt egttr> terrible bugbeare; <tt egttr> one that ouerthrowes
all he meets with; <tt usg> (Ironically.)