Part 1
Understanding embedded clauses
01What is an embedded clause?
When we speak, we often need to express not just simple actions, but thoughts about thoughts. Consider the difference:
Simple: She left. Embedded: I know that she left.
In the second sentence, "she left" is no longer a standalone statement. It has become the content of knowing, tucked inside a larger sentence. The embedded clause "that she left" is what I know, the object of my mental verb.
This lets us take a complete thought and make it part of a larger thought. We can report what others said, wonder about possibilities, express beliefs, and stack knowledge upon knowledge, all through embedding.
02Three types of embedding
Languages generally embed three kinds of content. The first is statements: facts, beliefs, reports of what someone communicated. I know that the sun is warm; she believes that rain will come; he said that you arrived. The second is questions: uncertainties, wonderings, inquiries being considered. I wonder whether she will come; she asked if I had seen the book; we don't know whether it's true. The third is quotation, someone's exact words preserved verbatim. She said, "I am leaving." He asked, "Are you well?" The teacher announced, "Class begins now."
English handles these differently. Statements use "that" (often optional). Questions use "whether" or "if." Quotations use quotation marks in writing, and intonation shifts in speech.
Phi handles all three with one system: paired complementizers that open and close the embedded material, like spoken brackets.
03How English does it
In English, embedding is somewhat casual. The word "that" introduces embedded statements, but speakers often drop it:
- I know (that) she left.
- He believes (that) it's true.
- She said (that) you called.
The boundary between the embedded clause and the main clause is unmarked at the end. English relies on word order and context to signal where the embedded content stops.
For questions, English uses "whether" or "if":
- I wonder whether she came.
- He asked if you were ready.
Again, no explicit closer. The sentence just ends, or the main clause continues without marking.
For quotations, English uses quotation marks in writing. In speech, there is no audible marker. A subtle pause or intonation shift might signal the quotation, but nothing explicit:
- She said, "I'm leaving."
- He announced, "The meeting is canceled."
This works reasonably well in simple sentences. But consider:
- I believe that she knows that he said that you think that it's true.
Where does each embedded clause end? English speakers parse this through accumulated practice, but the boundaries are implicit, recoverable only through semantic plausibility and verb valency.
04How Phi does it differently
Phi makes all boundaries explicit. Every embedded clause gets both an opener and a closer: spoken brackets.
| Type | Opener | Closer | Function |
|---|---|---|---|
| Declarative | tha | tho | Embeds statements (that...) |
| Interrogative | pha | pho | Embeds questions (whether...) |
| Quotative | sha | sho | Preserves exact words ("...") |
The opener announces what kind of content follows. The closer marks exactly where that content ends. The main verb then follows the closer, unambiguously.
English: I understand that they left. Phi structure: I [DECL.COMP they left DECL.COMP.CLOSE] understand.
mia tha shia to wepu tho shelomu
1SG DECL.COMP 3SG PST go DECL.COMP.CLOSE understand
The tha opens the embedded statement. The tho closes it. The main verb shelomu follows, and there is no ambiguity about what belongs to the embedded clause versus the main clause.
05Why closers matter
You might wonder: why does Phi need closers? English manages without them.
The answer lies in Phi's word order. Phi is strictly SOV (Subject-Object-Verb), meaning verbs come at the end of their clauses. When you embed one clause inside another, you get verb-verb sequences:
Without closer (ungrammatical):
*mia tha shia to wepu shelomu 1SG DECL.COMP 3SG PST go understand
Is this "I understand that they left"? Or is it attempting to say something about "they leaving-understanding"? Where does wepu end and shelomu begin? Two verbs in sequence, no boundary marked.
With closer (unambiguous):
mia tha shia to wepu tho shelomu
1SG DECL.COMP 3SG PST go DECL.COMP.CLOSE understand
Now the boundary is explicit. The tho says so: the embedded clause has ended. What follows is the main verb.
This is especially important when clauses nest:
mia tha thia tha shia to wepu tho phaelo tho shelomu
1SG DECL.COMP 2SG DECL.COMP 3SG PST go DECL.COMP.CLOSE feel DECL.COMP.CLOSE understand
Three verbs in sequence: wepu, phaelo, shelomu. Without the closers, parsing would be impossible. With them, each tho resolves exactly one tha, like matched parentheses in mathematics.
06The opener-closer pattern
Phi uses a consistent phonetic pattern for complementizer pairs:
- Openers end in
-a: tha, pha, sha - Closers end in
-o: tho, pho, sho
The vowel shift is not arbitrary decoration. a is the openest sound the mouth makes; o rounds the lips toward closure. Every time a frame opens or closes, the mouth acts out the boundary: jaw dropped at the opening, lips drawn round and nearly shut at the close. When you hear a complementizer ending in -a, something is opening. When you hear the same root ending in -o, it just closed.
07Cross-linguistic perspective
Phi's approach resembles several natural languages more than it resembles English:
Japanese uses quotative particles (to, tte) that follow quoted material, and clause-final particles that mark clause types. While Japanese does not use paired brackets exactly as Phi does, it shares the verb-final structure that motivates explicit boundary marking.
Korean similarly places verbs at the end and uses various complementizers and quotative markers to manage embedded clauses.
Turkish embeds clauses through nominalization and explicit case marking, which keeps the structural boundaries of its verb-final sentences clear.
What these languages share is the recognition that when verbs cluster at clause boundaries, explicit marking prevents ambiguity. Phi takes this principle further by systematically pairing every opener with a closer.
Lisp and formal languages offer another parallel. In Lisp programming, every opening parenthesis has a matching closing parenthesis: (outer (inner content) continues). Phi's complementizers work the same way: every opener finds its closer, and the whole parses without guesswork. This is not coincidental. Phi was designed with awareness that spoken language can be structurally unambiguous when delimiters match.
08Where this leads
The next section explains why the closers exist and why whu alone does without one. The three sections after it give each pair its own room, and the rest of the pamphlet turns the system into reflex: nesting, the errors English invites, exercises, and a reference page.
None of it is decoration. Every belief you report, every doubt you voice, and every word you repeat for someone absent will pass through one of these frames. A speaker who can open and close them without thinking is free to attend to what actually matters: whose thought is being held, and how carefully.