• ~× فضفِـض ×~

If we are dealing with speech, we need a phonological model of the word; if we are dealing with reading, we will need an orthographic one.

We can assume that the two are closely related, and are connected to the same unit of meaning. So forms of the word GIVE might be linked in this way:

• The situation is not quite as simple as this account suggests:
✓Both phonological and orthographic representations have to allow for variation → They have to take account of the fact that a speaker may have any one of a number of regional accents or that a written text may appear in any one of a number of different typefaces.
✓ There is the issue of homonymy. For example, there is a single phonological and orthographic representation of the word LIKE, but it must be connected to two entirely different entries, as in:
I LIKE ice-cream. She looks LIKE my sister.

• We also need to store information on the derivational morphology of a word.

• The meaning component of a lexical entry is sometimes referred to as its lemma.
• The lemma includes syntactic information about a word.