In addition to existing sentence structures and constructs, the syntax program can learn other common constructs. Although some may be understandable at present, others may require some additions to the program to fully understand. Below are some examples of other simple sentence constructs and how the program would interpret them.
"The man has a blue hat" (adjectives)
Here, the only unknown word with present knowledge is "blue". The program at present would interpret it using the word-context method, resulting in a type of "abnoun bnoun". Continuing in this manner would quickly lead to complications, however, so the program could be extended to understand words based on the form of types which are not nouns or verbs. The word directly before blue has the type "bnoun", so "blue" would be interpreted in this manner as a noun, resulting in two nouns in a row. The program could additionally be edited to interpret two like words in a row as a "noun phrase" or "verb phrase", which would differentiate adjectives from nouns and adverbs from verbs in a way more intuitive to the computer.
"The hat is blue" (adjectives as nouns)
This sentence presents a different use of an adjective effectively in the place of a noun. Assuming "is" had been previously learned as a verb, this sentence would be more-or-less readily understandable as "blue" would be interpreted as part of a noun phrase and hence related to nouns. This introduces another common sentence construct as well, namely that some nouns can appear directly after verbs without a "bnoun" word in between. Here, it may become worthwhile to add some indicator to new nouns about whether they can appear directly after a verb or not.
"The man has one hat" (numerical words)
This sentence would introduce numerical words, which would be readily understandable as "bnoun" words similar to "a" or "the". This is sufficient and accurate for most cases, though as the program expands into semantics this construct may require more specific definition.
"The man has four hats" (plurality)
Here, the concept of plurality is introduced. This sentence would simply be another example of known sentences with regards to syntax alone - the word "hats" would just be considered a new noun. However, this would likely be the most problematic concept regarding semantics. Without an external concept of meaning or some other indication, plurality would have to be learned simply by similarity at the word level. In many cases this would be sufficient, such as "hats", but some words do not follow standard plurality rules such as "mice" versus "mouse". Similar issues apply to verb tenses, though here the rules are even less standardized. This issue could be at least partially remedied by creating an artificial semblance of external understanding, though this would likely prove difficult as well.
Prepositions and Embedded Clauses
Many other sentence constructs are built from these sorts of basic patterns. For example, another common sentence construct involves prepositions such as the sentence "the man threw the hat in the trash". Assuming prior knowledge of the nouns present, the program could interpret "in" as verb-like, appearing between two noun phrases. This is, again, a sufficient interpretation in most cases. A marker indicating what would effectively be two full constructs could be implemented as well. Embedded clauses would be interpreted similarly, where the program would find the pattern of the outside clause and then the pattern of the inside clause. An understanding of punctuation would make such sentences more easily interpretable.