Language Should be Orderless
It's been quite a while but looks like I'm finally going to be able to explain this conlang that has been sitting in the back of my head for almost a year now. The core motivation behind this conlang is to accurately represent the relationships between arguments in a more naturalistic fashion than something like Discourse Representation Theory. I also hope it cleanly deals with some issues proposed by poststructuralist semiotics by grounding meaning in the real world. In a nutshell, this conlang outlines a manner in which all grammatical structures can be represented through a combination of suffixes and a linguistic particle that is closest to a 'verbified' preposition. It also may propose a different view of language than the native English speaker is acquainted with.
There are a number of ideas that seem unrelated that are part of this conlang, so I'll start with parts of speech. In English, the major categories are nouns, verbs, and adjectives. I don't think these are that useful of categories though, so words in this conlang will be classified as either statives or relationals. Statives and relationals both may be though of as similar syntactically to verbs here, but semantically, statives encompass what are typically thought of as nouns or adjectives in English. In this conlang, there isn't really a semantic difference between statives and relationals, the difference is rather a syntactic one. Statives take as argument one variable, while relationals take as argument multiple. For example, take the stative 'kutaka', which in English roughly means 'to be a dog'. If we say 'kutaka-ka', we pass the variable 'ka' to the stative 'kutaka', thereby making the proposition that 'ka' is a dog. The framework in which one thinks about this is different from English though. The stative 'kutaka' doesn't actually mean 'to be a dog', but rather 'to have the trait of dogness'. By the previous statement, we are saying that there exists a real-world object that has the trait of dogness.
Let's add some more words to the mix. The first person pronoun is a stative denoted by 'kapihi', so if I say 'kapihi-ku' we again state that there exists something that has the property of being myself. Introducing a relational, 'lika' which means something like 'to like' or 'to enjoy', one may now construct the sentence 'kutaka-ka kapihi-ku lika-ku-ka'. Relationals take two arguments, such as 'ku' and 'ka' here, and describes a directional relationship between them. The first argument is treated as the agent/subject and the second as the patient/object. Note that the distinction of words into 2 clearly definable categories also removes the semantic ambiguities of whether intransitive subjects are agents or patients. For example, the verb 'to sleep' which has vastly different semantic properties cross-linguistically due to the subjective choice made between whether one actively or passively sleeps, is clearly rendered as a stative in this conlang. One objection that may come to mind is signifying verb tense on a stative. This is not a problem however, but will be explained in a future blog post. Off the top of my head, the topics that remain to be covered on this conlang include: output flags for meta-marking which in turn covers tense/aspect/mood, adverbs, and verbal heads as objects; prepositions as relationals; two channel model with one for semantics and the other for syntax; high and low descriptivity statives; an in-depth analysis of this conlang's relation to historical work in semiotics.