Latin Dictionary Tools Page
Latin Dictionary Tools:
Philosophy. Since computers are used extensively, a dictionary as a tool to lookup for a word in it is not changed as dramatically as it would be expected, i. e. electronic dictionaries utilitarly more or less do the same work as dictionaries issued as books to do. The reason of it lays in the utilitar use of the paradigms to find eventually item somewhere from philologists point of view and to find it resultatively as response to a previous request from computer scientists point of view. The only innovation is the electron book, actually the paradigm is the same as in the case of the printed book. Thus, what we need is to acquire new praxis which would change the paradigm in general.
Dictionary's slight extention is form recognizer or speller. It very rearly has a printed book's equivalent. Here we are going to deel extensively just with form recognizers but behaving this way we aknowledge that this is done because of lack of idea how to widen the necessary paradigm of dictionaries in general and not because of seeing in it the end and our goal. Form recognizers are in wide use for contemporary languages but this is not the case for ancient ones, thus, Whitaker's form recognizer Words is a unique case; it would be good such have for a Greek too, but we do not know whether such a program exists. Commercial programs in these cases use annotated text which makes feel that forms are recognized online, but actually form recognition is made before preparing the text for use. Similarly behaves a large Perseus project. Thus one of our project main issues is form recognizers and how to build them. It must be mentioned that grammarians sometimes (or almost always) do not see difference between form recognizer and simple dictionary, criticizing former of not having all nice features what dictionaries have, actually not seeing that it is not possible principially. But form recognizers do much more than simple dictionaries. Form recognizer may define a language in an ideal case what simple dictionaries can not afford prinicipally. Actually, a form recognizer defines a language and it remains to find how precise this definition is or is only a rough approximation of the language it is aiming at. If some form recognizer seems to be a bad language definition then criticizers of the fact first must recognize that no dictionary ever may have some pretension to such a definition. Thus, developing of the science of the form recognizers has much more significance as we can predict in our negligence of the subject. By the way, one of acquisitions of this page is that we have possibility to show how Whitaker's speller [or our equivalent of Whitaker's speller or the program Words] fits as Latin definition and where and in what cases it can not do it well.
To start with we took Whitaker's form recognizer because we do not know of an existence of something equivalent. As it is our goal to build form recognizers by ourselves it does not matter what we take as a first sample. Either we do not need to build first sample by ourselves because any experience already acquired is appreciated by ourselves. The main issues of the Whitaker's Word program are 1) flexion table, 2) meaning table and 3) matching between stems and inflects. First and third should be critical for every form recognizer to be developed. The only thing which form recognizer shears with dictionary and that does not belong to it crucially is the meaning table. Flexion table essentially comprise all paradigms of the language. Thus, the main thing that resolves all is the matching, which mathematically speaking is a function from stems to flections. Thus, if word w consists of stem s and flexion f, then w= s+f or some imaginablematching function sets the necessary correspondence m:(s,f)->w and thus, the matching m we need is m: S,F->W where W are all words in language, S all stems and F all flections. Does such matching m with sensible sets W, S and F exist that may pretend to be named definition of the language? Does such m exist for Latin? Actually, Whitaker's Word in his way and our form recognizer in our own way try to answer just this question. What Whitaker's Word has achieved up to now we find may be called excellent.
This site in general is based on the database which via active server pages /asp/ is made accessible in read only mode to the browser of the INTERNET user. Three main tables of the database are general stem table of Latin words, general inflect table and general meaning table, which are based on the data taken (with little changes) from Whilliam Whitaker Words program and supported with necessary queries to make possible recognition of the Latin forms. The database contain also some closed dictionaries which do not possess power to recognize forms but only to list base forms of the Latin words. The main dictionary is unabridged Whitaker dictionary, others are subsets of this general dictionary. The database contain tools to produce such dictionaries automatically, i. e. using some program tools and a set of Latin forms, say, some Latin text even with word repetitions as inevitably it is expected to be. Of course, if some word is not present in full dictionary then followingly it would be absent in the subset dictionary too. It must be noted that following this approach meaning sets of particular dictionaries are subsets of the general meaning set which cases a side effect that in a very simple dictionary some simple word may have not as simple meaning list as it would be expected or appropriate when handmade.
Dictionary Tools present resources for several search functions:
The database in general comprise tools to recognize arbitrary Latin form. However, besides it contains some closed dictionaries which do not have power to recognize forms but only list words in their ground forms. Besides, dictionaries may be easy produced manually by some program tools in the database but it is not possible to do online. /Eventually the project is supposed to develope in a way to make this and similar to this functions possible online too./
A new dictionary may be produced from a list of Latin forms, e. g. there may be taken an arbitrary Latin text, and the new dictionary should be a subset of the general dictionary which should contain all words present both in the text and in the general dictionary where the meanings should be taken from the general dictionary.
Up to now there are five dictionaries present in the
database. These are:
How to use this search function in educational purposes?
Choose some dictionary and find, say, all nouns of the 5-th declension, or all verbs of 4-th conjugation, or rather more sofisticated search, say, all feminine nouns of 4-th declension, how many they should be in different dictionaries. Finding some word one may wish to see all the group of words similar to it, e. g. a noun tussis 'cough' belongs to a group of similar 7 nouns in the Whitaker dictionary /find what parameters this group possess, are they 3, 8?/. Look up this word in the flexion table generator. What specific distinct feature it has? Are there any of this group in the Wheelock dictionary?
Flexion table generator produces flexion table for an arbitrary Latin form entered in the data field of the table. If no other parameters are specified then some default parameters are supposed. Better way to find out how the generator works is to try several steps and find it out experimentally. Nevertheless some suggestions would be useful. First please try a form without any specifying parameters. For example, if you want to see flexion table 'amo, amas, amat...' then give any form of this list or even infinitive 'amare' . If you want to have subjunctive imperfect table for the form you have chosen then corresponding parameters must be specified, i. e. Verbi Tempus: IMPF, Verbi Modus: SUB . The only thing that must be taken into account that parameters must be respecified in every step because they are not remembered by server. If you give some form and in return receive a table you did not want then you must specify additional parameter which leads you to the presupposed result. The generator is built supposing that it is always possible. Please, announce if in some case you discover that it is not possible.
This is only a test version of the form generator. Thus, some words return not quite a good table, i.e. there may be more values then supposed by the table of the forms /paralel forms may appear which the generator can't display regularly/, or even more unpleasant irregularities may appear.