Correlatives

Proposal 1

Naturalistic, with each word being taken from the corresponding word in the source languages.

Advantages

• Easier to guess the meaning of a word

Disadvantages

• Harder to learn, as there are more words to be memorised

Proposal 2

Schematic, with each correlative being predictable from its meaning by a combination of two words, along the lines of Esperanto's system. For example, all of the interrogatives would start with the word vilk, and all correlatives about location end with sted, so where would be vilk sted or literally which place.

Advantages

• Perhaps easier to learn, as there are fewer words to be memorised

Disadvantages

• Somewhat unintuitive as this is less naturalistic and less reflective of what the reference langs do

Proposal 3

Schematic, with all of the correlatives predicable from their function, but more reflective of the reference langs. This is possible due to some schematic-ness already existing for historical reasons. For example, in English, most interrogatives start wh- and most definites start with th-. Some locatives end in -ere, and some correlatives about things end in -at. This accurately predicts "what", "that", "where", and "there" (at least orthographically). A similar pattern exists in other Germanic languages.

Advantages

• Perhaps the best of both worlds - easier to learn and easier to recognise

Disadvantages

• The system is not perfectly followed in the reference languages ("here" and "this" cannot be predicted, for example), so there would still need to be some degree of irregularity and/or change from the reference languages.