I am wondering about conlangs and thinking about English currently. I'm wondering does English have a finite set of patterns for constructing sentences? That is, could you build a computer program that could perhaps combine and recursively expand some more primitive patterns so as to give you a clear set of rules for creating sentences? Or are there no such patterns and anything goes... eventually?

Some examples of things that aren't "proper" sentences:

Store to the go I want to. (Said in a skip-step fashion)
If we store go later miss the train we will perhaps.

We can figure out roughy what these mean even though from my knowledge I didn't have a schema in my head defining this pattern. But maybe it's composed out of simpler pattern rules I learned along the way.

I wonder because at first I was trying to construct a conlang with clear rules for verbs, adjectives, nouns, and adpositions ("relations"). By adding special endings to words perhaps. But this results in longer words and thus longer sentences than you have in English. So I was trying to figure out how English works in this sense, how they are able to just use the sentence morphology (word order and such, syntax?) to get the meaning across, without necessarily word endings. It seems that to come up with such a system like English, you would have to define every pattern in advance. But that's unrealistic, so somehow it evolves without being defined in advance, but I can't figure it out yet, how English did it.

Part of the reason I ask is because everyone keeps saying English generally has a strict word order. But what is this order? How did it get established in the beginning if you could theorize? That's where I'm coming from. If I can figure out how to limit word order in a specific set of patterns, then I can come up with a con-lang like English. But I don't see how to set up the initial patterns. In order to do it as a conlang, you would have to know all the ways (or many ways) to model information and convey information like a computer. There are actions, and objects, etc.. You basically are a master at philosophy. Then you see how you can streamline those models to sentences/linear-sequences so they can all interact. There would have to be easily hundreds of them. But I think one person can do it, figure it all out... It's just that I don't see it yet, how to convert generic computer-like knowledge/models of semantic structures and processes into sentences specifically.

  • 1
    They say it has a strict word order because violations of the order are instantly recognizable by native speakers. They don't say it order strict has word a.
    – jlawler
    Commented Dec 26, 2020 at 0:13

2 Answers 2


In one interpretation of the concept "pattern", there are infinitely many patterns. Some example of "patterns": [det N aux V det N], [det adj N aux V det N], [det N aux V det adj N], [det N aux V det N V], [det N aux V det N V det N]. (Examples: the boy has seen the girl, the little boy has seen the girl, the boy has seen the little girl, the boy has seen the girl cook, the boy has seen the girl cook the rice). The field of theoretical linguistics has developed any tools for reducing the infinity of patterns to a finitude of rules that generate those patterns. You could write a computer program to grind out sentences: people have done so.


Well, yes, it is finite. It would only rise to infinite if you have a sentence with infinitely many words. And Computational Linguistics is an entire field based around going off of viable patterns in order to construct sentences with an AI.

English is Adjective-Frontal SVO, and it is only as such because that's what it became over its development. For example, Latin was SOV, but Spanish, French, Italian and such are SOV. Changes simply tend to occur, typically only in certain patterns.

SOV is the most common type, and theories believe it to be the most natural form in a sense. This can change to OVS, SVO, or OVS. and SVO can go to VSO or VOS. And this triangle tends to change between each other.


I would recommend looking into Head-Marking and such: https://en.wikipedia.org/wiki/Head_(linguistics)


These help a lot in understanding the way that language is structured.


Language is based upon the order of Subject, Object, and Verb, in the largest sense. Then one must include where the Articles, Markers, and such go. These typically work forward towards bringing about certain features.

For example, if the object markers go behind the modified noun, they would, if they do combine, would combine as a suffix. But if they were initial, it would combine as a prefix. This is how affixing operates, through modifiers and nouns, overtime, molding together, and the manner in which this occurs is dependent upon the placement of the modifier. This combination does not always occur, however, and in times it could occur with certain modifiers but not others.

In times different modifiers can go on different sides of the modified noun, for example, the descriptor Rojo or Red, in Spanish would go after Casa, or House, making La Casa Rojo, but a numerical designator would go before, Dos Casas. But they could also agree, like in English, "The Red House" and "Two Houses". In short, it can all differ, but it usually does remain consistent within itself, you must simply learn the ways in which it remains consistent.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.