Syntax Tree Generator | How Language Builds Meaning

The foundation of language is the idea that meaning is made by words in combination with each other. We add words to other words to explain our world. The tool on this page is a syntax tree generator to help you visualize this abstract combination. For example, the word “Persian” in isolation does not carry much weight, but if you combine it with the word “cat” to make “Persian cat,” you have made something with more meaning than the two words in isolation. If we add another word, we can then take “Persian cat” and combine it with “fuzzy” to get “fuzzy Persian cat,” something with even more meaning

Everything we add to our sentence sequence gives more information to an internal representation, which new words are then added to. Linguist Noam Chomsky called this Merge and defined it as the foundation of human grammar.

What is merge?

Merge is an operation that combines two things at a time, never more than that, and what it produces then becomes the input for the next combination. This is, by mathematical definition, recursive; each step builds on the last and does so in a structured manner. That is how something as complex as language grows from something so simple.

Watch it accumulate. “Cat” merges with “Persian” to give “Persian cat.” That joins with “fuzzy” to give “fuzzy Persian cat,” and that combines with “rumpled” to give “rumpled fuzzy Persian cat,” and that finally goes with “the” to give “the rumpled fuzzy Persian cat.” Separately, “is” merges with “sleeping” to give “is sleeping.” Those two finished pieces then combine with each other: “The rumpled fuzzy Persian cat is sleeping.” To place the cat, we build “the sun” from “the” and “sun,” merge that with “in” to get “in the sun,” and finally the end result is the sentence we already had, and we end with “The rumpled fuzzy Persian cat is sleeping in the sun.” Our ending sentence is worlds of complexity away from the two words we started with.

We can illustrate the above complexity pictorially. Interestingly, the way our words came together is different from the way we speak or write them. This tool, or syntax tree, is a diagram of how words combine in meaning, presented as a pictorial flow of information, added step by step.

When we look at a syntactic tree, we see the output as an illustration of Merge. The bottom is where we start, in our example, “cat” and “Persian,” and it then ends at the top with our finished sentence. As in any language, the order of words and the order of Merge produce different meanings.

For example, “The dog bit the man” vs. “The man bit the dog” produce sentences with completely different actions based on word order, and “only she saw the cat” vs. “she only saw the cat” vs. “she saw only the cat” create different meanings based on different orders of Merge. Look at the word “only” and see what it joins with: only she, only saw, and only the cat. In the first sentence, she was the only one who saw the cat; no one else did. In the second sentence, she only saw the cat, but physically did nothing else. Finally, in the last sentence, the cat was the only thing she saw. We create endless meaning with combinations.

Chomsky argues that this complexity of Merge is what makes human language unique because we take a finite vocabulary and combine it to create infinite possible sentences; there is no end to adding units together to generate new meanings. It is possible to think that Merge began when early man started to move from utterances that carried multiple meanings to those that needed to be combined to carry more information. Music and rhythmic utterances in particular would lend themselves to early protolanguage development.

This syntax tree generator tool was made to illustrate and simplify the concept of Merge. Try creating a sequence of your own. The meaning does not matter as much as the process.

How to Use the Syntactic Merge Lab

Click words in the lexicon panel on the left. They appear in the workspace on the right. Hover any word to see the features it carries.

Click two objects in the workspace to select them, then press the Merge button. The button shows you what the merge will produce before you commit.

Build outward from the smallest pieces. A determiner like the is not a complete phrase until it has merged with a noun, so merge the with cat first, then merge that result with the verb.

Merge never refuses. If two things cannot legally combine, the lab builds the object anyway, marks it in red, and explains what went wrong. That is the point, rather than a malfunction: Merge is free, and grammar is the separate step that filters what Merge produces.

Under the tree, a verdict tells you whether your derivation converges, and lists what is still missing if it does not.

To move a constituent, select a whole object, then click a piece inside its tree. That is Internal Merge, the operation behind questions. Try building Ø the cat chases what, then moving what to the front.

The tabs:

• Tree shows the structure as it builds
• Sets show the same structure as pure set notation, plus labeled brackets
• Algebra covers the mathematical properties of Merge
• Census builds every possible bracketing of your words, checks each one, and reports how many survive

To add your own words, type into the Custom field, choose a category, and press Add. Verbs let you choose transitive or intransitive; determiners can be marked as movable question words.