> For the complete documentation index, see [llms.txt](https://docs.lpp-minduniverse.org/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lpp-minduniverse.org/lingua-pactum-protocol-lpp-documentation/10.-research-and-publications/10.1-research-program-overview.md).

# 10.1 Research Program Overview

The LPP research program began with a conceptual problem:

> **Runtime permission does not establish the legitimacy of the action being permitted.**

That observation led to the definition of a governance layer upstream of ordinary execution authorization.

Subsequent papers progressively answered the implementation questions created by that initial position.

#### Stage 1 — Establish the Layer

Paper 1 asks:

> Why is ordinary runtime governance insufficient?

It introduces **constitutional admissibility** and establishes Layer 0 as distinct from policy enforcement and execution authorization.

#### Stage 2 — Operationalize the Layer

Once Layer 0 exists conceptually, the next question is:

> How can the system compute an admissibility result?

Paper 2 introduces the **LPP Admission Kernel**, formal Admission Requests, authority state, lifecycle, failure semantics, Execution Permits, and the Layer 0 / Layer 1 interface.

#### Stage 3 — Test the Model Against a Consequential Domain

A general protocol is not enough if it cannot be translated into real execution conditions.

Paper 3 applies the model to financial AI.

It asks:

> What does constitutional admissibility mean when the action can move assets or create financial consequence?

This produces LPP-FIN.

#### Stage 4 — Formalize the Legitimacy Carrier

The earlier papers rely on authority.

That creates another question:

> What exactly is the machine-verifiable object representing that authority?

Paper 4 defines the **Authority Object** as a structured, scoped, lifecycle-bound, revocable, and verifiable legitimacy carrier.

#### Stage 5 — Govern Meaning Across Time

A valid Authority Object still leaves a remaining problem.

An AI trajectory may begin under legitimate authority and later change meaning.

Paper 5 therefore asks:

> How can an initially authorized trajectory become inadmissible without the original Authority Object itself becoming invalid?

This leads to:

* Human-Authorized Intent,
* latent intent mismatch,
* material mutation,
* Semantic Verification,
* trajectory-level admissibility,
* mandatory re-admission,
* and SVNN verification evidence.

#### Research Program as a Dependency Chain

The resulting program should therefore be read as:

```
Paper 1:
A Layer 0 is necessary.

        ↓

Paper 2:
Layer 0 can be expressed as an Admission Kernel.

        ↓

Paper 3:
The Admission Kernel can be instantiated in a consequential domain.

        ↓

Paper 4:
Admission requires an explicit machine-verifiable legitimacy object.

        ↓

Paper 5:
Valid authority must also remain semantically aligned
with the evolving execution trajectory.
```

This dependency structure is more important than publication order alone.

***