CSGI
CAPABILITY DEVELOPMENT PROGRAMME
Demo build
Priya Mehta Deliberating member
Active
TG Hub CHI Counsel UGC Unicator
Prototypes
CSGF Plenum
CSGF Academy
← Return to CSGI
Action required. The CSGI Sentinel has flagged that A3 – The limits of what we can say – is a prerequisite for your current decision assignment. Complete it before the deliberation stage opens on 14 March.
1
Required now
1
In progress
2
Completed
8
Remaining
A3Required – due 14 March
The limits of what we can say
Linguistic and representational constraints – how governance fails at the level of language, category, and symbolic compression before institutions are even involved.
~75 min – prereqs metBegin programme →
A2In progress
The limits of what we can see
Cognitive constraints – bounded attention, non-linear complexity, and the structural simplification every governance actor is forced to make.
60% – guided engagementResume →
A4Unlocks after A2 + A3
The limits of shared meaning
Cultural and interpretive constraints – why shared vocabulary does not produce shared understanding, and how this shapes governance before rules are applied.
~80 min – 2 prereqs outstandingLocked
A1 – Why governance fails
Completed 18 February 2025
View record →
B-DM – Deliberating member induction
Completed 12 February 2025 – CSGI role unlocked
View record →
TRACK A – TIER 1: FOUNDATION (all roles)
A1 – FOUNDATION
Why governance fails
Structural overview of the five constraint layers and their interaction. The conceptual map for the full programme sequence.
~45 minComplete
A2 – FOUNDATION
The limits of what we can see
Cognitive constraints – bounded attention, non-linear complexity, forced simplification and its costs.
~70 minIn progress
A3 – FOUNDATION
The limits of what we can say
Linguistic and representational constraints – semantic fragmentation and category collapse.
~75 minRequired now
A4 – FOUNDATION
The limits of shared meaning
Cultural and interpretive constraints – interpretive drift, symbol substitution, normative fragmentation.
~80 minLocked
A5 – FOUNDATION
The biological layer
Human nature constraints – loss aversion, tribal heuristics, temporal bias as structural governance constraints.
~85 minLocked
TRACK A – TIER 2: INSTITUTIONAL (all roles)
A6 – INSTITUTIONAL
How institutions embed distortion
Five mechanisms: accountability fragmentation, procedural inertia, incentive encoding, recursive entrenchment, decision integrity gaps.
~90 minLocked
A7 – INSTITUTIONAL
Decision integrity – from principle to practice
What decision integrity means operationally and how the CSGI lifecycle addresses it.
~80 minLocked
TRACK A – TIER 3: ROLE-SPECIFIC
A8 – MANDATE AUTHORITY
Mandate and incentive design
How the authority setting mandates builds in or prevents institutional distortion patterns.
~75 minNot your role
A9 – DELIBERATING MEMBER
Deliberation under constraint
How cognitive, cultural, and behavioural constraints operate specifically in deliberative settings.
~80 minLocked
A10 – OBSERVER / AUDITOR
Observation and pattern recognition
How to read a governance system for embedded distortion and accumulated decision drift.
~85 minNot your role
A11 – RULES COMMITTEE
Designing for coherence
All five coherence capabilities as design problems for governance architecture.
~100 minNot your role
TRACK B – IN-APP ONBOARDING (inside CSGI)
B-DM – TRACK B
Deliberating member induction
System-specific onboarding completed inside CSGI. Unlocks full deliberating member permissions.
~30 minComplete
2
Programmes completed
1
In progress
1
Required now
CSGI SENTINEL VIEW
Role eligibilityEligible – deliberating member
Active flagA3 prerequisite – due 14 March
B-DM inductionComplete – 12 February 2025
Last updated18 February 2025
ProgrammeCompletedDeterminationDurationCertificate
A1 – Why governance fails18 Feb 2025Complete42 minDownload
B-DM – Deliberating member induction12 Feb 2025Complete28 minDownload
A2 – The limits of what we can seeIn progress
A3 – The limits of what we can sayRequired now
TRACK A – FOUNDATION – TIER 1
The limits of what we can say
Programme A3 of 11 – follows A2: The limits of what we can see
Governance does not act on reality directly. Every decision, standard, mandate, and procedure reaches the world through language – through categories, definitions, and terms that compress complex situations into forms that can be communicated and administered. That compression is unavoidable. But it introduces distortion at the very point where reality becomes governable.
In programme A2 you examined the limits of what the human mind can perceive and track. This programme moves one step further along the causation chain: even when perception is adequate, the act of translating what we perceive into language introduces a second layer of structural failure. The terms governance systems depend on – accountability, risk, responsibility, public good – do not hold stable meaning across the actors, institutions, and time horizons through which they must operate.

Continuing from A2. Programme A2 established that cognitive limits force every governance actor to compress complex systems into partial models. This programme examines what happens next: that compression must be expressed in language – and language introduces its own structural distortion before any institution or rule is applied. The two constraints are sequential and compounding.

This programme examines two specific failure modes – semantic fragmentation and category collapse – and traces how representational distortion compounds as language moves through institutions. By the end of the guided engagement you will be able to recognize both failure modes in live governance contexts and understand why they precede, and are not caused by, the institutional structures that eventually make them visible.
Your progress is saved automatically.
You can exit and return at any point.
LEARNING OUTCOMES
Distinguish semantic fragmentation from category collapse and identify each in governance documentation
Explain how representational distortion compounds through institutional transmission
Recognize agreement that is terminological rather than substantive in deliberative settings
Connect representational failure to the problem framing stage of the CSGI lifecycle
PROGRAMME DETAILS
Duration~75 min
TrackA – Foundation
Tier1 of 2
Prerequisites 
A1A2
CONTENT AUTHOR
TG
Terravive Group
PROGRAMME CONCEPTS
The representation problem
Covered
Semantic fragmentation
Covered
Category collapse
In progress
Transmission and compounding
Not yet reached
Recognizing it in practice
Not yet reached
Connection to CSGI problem framing
Not yet reached
CDT
Capability Development Tutor

Let's start with something concrete. You are a deliberating member on a decision about infrastructure maintenance prioritization. The mandate authority has issued guidance that all decisions must account for public safety risk.

Before we go any further – what does that term mean to you? Not the official definition. What does "public safety risk" actually refer to when you use it in a governance context?

PM
Priya Mehta

I would say it means the likelihood that a failure in infrastructure could result in harm to members of the public – so probability of failure combined with the severity of the outcome if it occurs.

CDT
Capability Development Tutor

That's a coherent definition – probability times severity. It's also how an engineer would tend to define it. Now consider that the finance officer on the same decision is using the same term to mean something closer to legal liability exposure: the risk that the institution becomes accountable for harm it failed to prevent.

Both of you are using "public safety risk" correctly. Both definitions are legitimate. But they point to different things – which means they would produce different prioritization decisions if applied consistently.

This is category collapse: a single label compressing genuinely different referents. The term appears to create shared ground. It doesn't. What do you think happens in practice when a deliberating body proceeds on the assumption that the term is shared?

PM
Priya Mehta

I suppose people end up talking past each other without realizing it. They think they've agreed on the problem but they're actually solving different problems. You'd only discover it when the options being proposed don't make sense to someone else.

CDT
Capability Development Tutor

Exactly – and notice when the divergence becomes visible. Not during problem framing, where it could be addressed. Not during options analysis. It surfaces during deliberation, when people are already committed to positions. At that point the disagreement looks procedural or political. It gets managed as conflict rather than resolved as definitional confusion.

This has a specific implication for your role. As a deliberating member, you are operating at the stage where category collapse is most likely to already be embedded in the framing you have been handed. The question is whether you can recognize it when it's there.

Let me give you a harder case. In the same decision, the term maintenance backlog appears in the options analysis. What are the different things that term could plausibly mean to different members of this deliberating body?

Progress
35%
This is a guided engagement – the programme will move through all concepts before proceeding to stage 3.
STAGE 3 OF 4 – WRITTEN RESPONSE
SCENARIO
Infrastructure maintenance prioritization – Rajasthan State Roads Authority

A deliberating body has been convened to decide which of three infrastructure maintenance programmes should receive priority funding allocation for the coming fiscal year. The mandate authority has framed the decision around two criteria: public safety risk and maintenance backlog severity.

During the options analysis stage, two members of the body have submitted conflicting assessments of Programme B. One rates it high priority on both criteria; the other rates it low priority on both. When asked to explain, both cite the same data set and the same mandate guidance. Neither believes they have made an error.

The deliberation stage is about to open. You are a deliberating member who has reviewed both assessments.

Your response
Drawing on what you covered in the guided engagement, explain what you think is happening in this scenario and what – if anything – you would do before the deliberation opens.
First submission. After you submit, the Tutor will provide feedback. You may revise your response once before a final determination is made.
0 words
150 – 250 words
Your response will be evaluated against the assessment criteria.
One revision is permitted after feedback is returned.
ASSESSMENT CRITERIA
Identifies the representational failure mode at work – category collapse, semantic fragmentation, or both
Explains why the divergence is structural rather than a matter of one assessor being wrong
Proposes a response appropriate to a deliberating member's role and the stage of the process
Demonstrates understanding that the failure precedes the deliberation – not caused by it
CONCEPTS COVERED
The representation problemSemantic fragmentationCategory collapseTransmission and compoundingRecognizing it in practiceConnection to CSGI problem framing
HOW THIS IS ASSESSED
Your response is evaluated by the Tutor against the criteria above. The assessment looks for evidence of conceptual understanding applied to the scenario – not recall of definitions. A brief summary of your response and the determination will be recorded in your competence record.
PROGRAMME COMPLETE
A3 – The limits of what we can say
Completed 6 March 2025 – first submission – Priya Mehta
ASSESSMENT SUMMARY

Your response correctly identified category collapse as the operative failure mode and located it at the representational layer – prior to any procedural or institutional cause. You distinguished this from a factual error on either assessor's part, which is the central insight the scenario was designed to test.

Your proposed response – raising definitional clarification before the deliberation opens – is appropriate to your role and to the stage of the process. It demonstrates understanding that the problem framing stage is the right moment to surface this kind of divergence, not the deliberation itself.

One area for continued attention: the response touched on semantic fragmentation but did not fully distinguish it from category collapse. Both are present in the scenario. This distinction becomes more important in programme A4, where cultural interpretation introduces a third layer of the same compounding pattern.

RECORD UPDATED
ProgrammeA3 – The limits of what we can say
DeterminationComplete – first submission
Date6 March 2025
Duration68 minutes
TrackA – Foundation, Tier 1
Competence recordUpdated
CSGI SentinelNotified
Sentinel flag cleared. The prerequisite flag on your current decision assignment has been resolved. The deliberation stage will be available to you when it opens on 14 March.
UP NEXT
A4 – FOUNDATION – TIER 1
The limits of shared meaning
Cultural and interpretive constraints – why shared vocabulary does not produce shared understanding, and how interpretive divergence compounds the representational failures examined in A3.
Begin programme →
Your record and CSGI status have been updated.
CERTIFICATE
A3
The limits of what we can say
6 March 2025 – Priya Mehta
YOUR SEQUENCE
A1 – Why governance fails
A2 – The limits of what we can see
A3 – The limits of what we can say
A4 – The limits of shared meaning
A5 – The biological layer
A6 – How institutions embed distortion
CDP Portal
Capability Development Programme