TL;DR: I'm trying to make a language semantically like Ithkuil (or better) but practical as an everyday language. To illustrate, I reformulate its Configurations, which you can see at the bottom of the post.

Motivation

One of the original inspirations for my language was reading about Ithkuil morphology, though not in a way necessarily favorable to it. There are a few clashes with the kind of language I want:

• The morphology is extremely complicated.
• The morphosemantics require certain aspects of a situational to always be specified.
• The semantics has many redundancies that (to my understanding) are there because there is no system to apply morphological categories in any way other than all at once.

All of these points are related because of Ithkuil's synthetic approach: (most) nuances come from inflections. I'd imagine this comes from the creator's goal of extreme semantic compression. However, I feel that the ability (not requirement) to specify many of these nuances in a regular way greatly contributes to the power of a language and should be done in a way that is more user-friendly.

Let's look at another way to handle these things, without getting into si-ka's grammar.

A quick tour of Ithkuil Configuration

Ithkuil specifies 9 Configurations, which I'll summarize with quotes from the official grammar:

• (UNI) Uniplex: "a single, contextual embodiment of the stem concept"
• (DPX) Duplex: "a related binary set"; "can […] be used to describe any set of two identical or complementary objects"
• (DCT) Discrete: "a grouping or set of the basic stem units that are more or less identiform"
• (AGG) Aggregative: "an associated group or set of entities, except […] the members of the configurational set are not identical to one another"
• (SEG) Segmentative: "a grouping or set of the basic stem units, the individual members of which are physically similar or identical and are […] so that the group moves or operates together"
• (CPN) Componential: "operates identically to the SEGMENTATIVE above, except that the individual members of the configurational set are not physically similar"
• (COH) Coherent: "functions similarly to the SEGMENTATIVE above, except that the individual members of the configuration are connected, fused or mixed with one another to form a coherent emergent entity"
• (CST) Composite: "the same as the COHERENT above except that the individual members of the configurational set are not identical or physically similar"
• (MLT) Multiform: "an individual member of a “fuzzy” set"

Combining similar concepts

Ignoring MLT for a moment, there are a number of similarities between these configurations:

• UNI entails that there is one of something, DPX entails two, and rest entail many.
• DPX, DCT, SEG, and COH entail the items are similar (as well as UNI, technically), while AGG, CPN, CST and MLT entail the opposite. In this respect, DCT-AGG, SEG-CPN, and COH-CST pair off.
• DCT-AGG (if either applies) = many related objects.
• SEG-CPN = many objects moving as a group.
• COH-CST = many objects that are fused together.
• MLT = many objects that are like what the root word means.

Now let's say we have a language where modifier words come after what they modifiy, and some modifiers

• 1: one of _
• 2: two of _
• many: many of _
• liquid: _ moving as a connected mass (I chose "liquid" because of liquids' microscopic properties; don't think of this as a physical state)
• solid: _ moving as a solid (fused) mass
• same: a _ that is the same (or significantly similar) in each case
• different: a _ that is different in each case
• like: something sort of like _

then we can describe all the configurations like so:

Simplification

• UNI: 1
• DPX: same 2
• DCT: same many
• AGG: different many
• SEG: same many liquid
• CPN: different many liquid
• COH: same many solid
• CST: different many solid
• MLT: like many

If all these modifiers are actually short (~2 phonemes each), this makes for a viable alternative in most situations, and we can also convey more nuances without resorting to some other mechanism (for instance the concept of many without asserting similarity or dissimilarity). Here's another version using ideographs to paint a picture:

一   同二   同多   別多   同多液   別多液   同多固   別多固   型多