This is my analysis:
There is a many-one relationship between nouns (words) and languages. A language has many nouns, but each noun belongs to only one language.
There is a many-many relationship between nouns and meanings of nouns. A word can have many meanings, while a single meaning can be expressed with different words.
This is what I propose for the design of the tables:
Table: languages
language_id
name
Table: nouns
noun_id
language_id
noun
Table: meanings
meaning_id
meaning
Table: nouns_meanings
id
noun_id
meaning_id
Consider two languages, English and Urdu. According to a dictionary I checked, the Urdu words for father are baaba, baap, pidar and waalid, and the words for mother are maaƱ and waalidah. Suppose that the English word 'father' is also considered to have the meaning 'priest'.
In the languages table:
1, 'English'
2, 'Urdu'
In the nouns table:
1, 1, 'father'
2, 1, 'mother'
3, 1, 'priest'
4, 2, 'baaba'
5, 2, 'baap'
6, 2, 'pidar'
7, 2, 'waalid'
8, 2, 'maaƱ'
9, 2, 'waalidah'
In the meanings table:
1, 'father'
2, 'mother'
3, 'priest'
In the nouns_meanings table:
1, 1, 1
2, 1, 3
3, 2, 2
4, 3, 3
5, 4, 1
6, 5, 1
7, 6, 1
8, 7, 1
9, 8, 2
10,9, 2
So, if the user wants to add support for Spanish, he/she will add Spanish to the languages table, add the Spanish nouns to the nouns table, then add the relationships in the nouns_meanings table. If a Spanish noun with a meaning not listed is added, then the meanings table will also have a new entry.