# 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.