Latin Dictionary Tools in
/Instructions for use of the morphological database of Latin in
Dainis Zeps
better to use link below
Presented as a report in HLT conference held in Riga 21.-22.04.04.
Many ancient language learning tools have come in
use in Internet during last less then ten years:
of Classical Languages:
Greek, Latin;
of Biblical Languages,
Hebrew, Aramaic, Greek.
Most of all, Latin websites are produced, e.g.
• Latin Teaching Materials at Saint Louis University
– whole spectrum of learning tools
• Perseus Project: Texts and Translations / Latin and Greek /
– Morphological Analysis as tagging and Dictionary tools
• The Classics Technology Center: CTCWeb
– ample collection of Latin and Greek sites
• Reading Exercises in Latin /D. Zeps/
– morphological and lexical tagging, syntactical phrases’ tagging
What is electronic dictionary?
Most of all we find in Internet dictionary as a text
window supplied with push button
Push to find
Pushing the button dictionary entry opens with
corresponding headword.
In Perseus Project: Texts and Translations you would find:
In Perseus Project: Texts and Translations one
would find more:
...and even more
Thus, what concerns dictionaries, up to now mostly sophisticated
approach in ancient languages seems in
Perseus Project: Texts and Translations
Our aim is to suggest more sophisticated approach to treat term ‘dictionary’:
giving wider functionality to it.
Dictionary should be morphological tool for lexical items both with
analyzing and generating capability
Having full database relational capability;
Posessing easy of inquiring.
Last two features are contradicting: thus we suggest: full database relational
capability is that what should be integrated in the database on the whole, whereas
for user in Internet some chosen functions must be suggested that may be
gradually augmented according requirements of users.
Thus our aim is:
To use this dictionary tools page to get new experience in order to invent new
useful and mostly necessary functions of inquiring.
Another possibility would be try to exibite in Internet as many as possible
functions from database functionality, but it would make the use of the site too
complicated for unexperienced users.
Five sets of functions are presented:
better to use link below
Search Latin Form
Search Dictionaries
Flexion Table Generator
Latin Word Quiz
Multiplication Table Quiz
Searching Dictionary Tools
Please, use link
In the text window enter
Try ‘silvanus’, ‘amavisti’,
‘murilegus’, ‘astra’,
In the text window enter ‘slaughter’.
Try ‘father’, ‘fishmonger’, etc.
Push button Go. Change
number of subdictionary
/from 1 do 5/ in text window
and go on pushing Go.
Choose Dictionary and Search
Full dictionary and five subdictionaries
may be chosen.
Please, use link
Five subdictionaries are present which are possible to search.
Please try yourselves parameters as in picture and below:
dictionary ‘classic middle (Wheelock)’, word class ‘N’, Class ‘4’ and Genus ‘F’.
Please, use link
An example: inflection of the verb ‘amare’ in imperfect subjunctive:
please try yourselves: print ‘amo’ with verbi tempus ‘impf’ and verbi modus ‘sub’.
Exercises: please try:
‘bonus’ with gradus comparationis ‘comp’
‘prosum’ with pars orationis ‘v’ and verbi modus ‘sub’
‘duo’ with numeri genus ‘ord’
What else as some necessary
function could be added?
Exempli gratia,
In place of the named query parameters changeable parameters
may be added, i. e. above choise window stands another choise
window where we choose a function for the window below.
Representation of Latin form in
Set of Latin forms as morpheme
Whenever possible we stand to a simple rool:
A single Latin form = root + inflection.
Not always one morpheme in all its possible forms can be represented as
a root+inflection;
thus it is convenient to divide morpheme as if in submorphemes where
for each submorpheme there holds:
Submorpheme =root + inflection
E.g. morpheme for a Latin word ‘capere’ is set of submorphemes :
First present stem s.m. restricted on indicative allomorphs:
‘capi’ + {‘o’, ‘unt’}
Second present stem s.m. restricted on infinitive and indicative allomorphs:
‘cap’ + {‘ere’; ‘s’, ‘t’, ‘mus’, ‘tis’}
Perfect stem s.m. restricted on indicative allomorphs:
‘cep’ + {‘i’, ‘isti’, ‘it’, ‘imus’, ‘istis’, ‘erunt’}
Supine stem s.m. restricted on perfect allomorphs:
‘capt’ + {‘us’, ‘i’, ‘o’, ‘um’, ‘o’, ‘i’, ‘orum’, ‘is’, ‘os’, ‘is’}
More general question in a wider context in
AI in general: What may be our aim in
developing teaching tools in ancient
languages or in languages on the whole?
Two general approaches:
to make self learning systems to:
 to help us to learn (ancient) languages or
 to serve us as inquiring resources for getting
answers whenever we need something to know.
It may seem that both possibilities are crucial for
us as desirable outcome.
To answer this question we
suggest to have a look back in
Hale’s reading machine
Hale W.G. 1887. The Art of Reading Latin: How to Teach It.
Cornell University, Boston, Ginn & Co., pp. 31.
William Gardner Hale in 1887 suggested a method for learning to read Latin
and Greek texts. His method may be called Hale’s reading machine.
Hale’s reading machine: reading sentence, for each new
word all possibilities of coming syntactical construction
must be explicitly named and characterized.
Thus, for every word which by reading comes in our mind, all possible
syntactical functions of it in the sentence must be explicitly characterized.
Hale argued that we must learn a language to this level to acquire such
capability of understanding of the language that is the only way and escape if
only we realy want to have hope to read ancient authors fluently.
Hale argued that we may learn a
language to such a level that we
can acquire the capability of
understanding the language
accordingly his method in
simulating his reading machine
that is the only way and escape
if only we realy want to have
hope to read ancient authors
It was in 1887. How it is now?
What shows up contemporary praxis concerning this general question?
What approach is more preferable for mathematical linguists?
Mathematical linguists are getting inside essence of languages more and more
But... What way they go when questioned from either one or other point of
What they do objectively making self learning tools and having their own
experience in understanding languages both from outside and inside?
t.l.=language teaching linquists
m.l.=mathematical linguists
2nd op.
1st op.=linguists objectively representing first
2nd op.=linguists objectively representing second
Use link
Use link
Thank you for attention!

Slide 1