NOUS COGNITIVE INTELLIGENCE PLATFORM
Longitudinal Cognitive Signature Analysis
This analysis presents the first longitudinal comparison in the NOUS case study series — the same individual measured across two distinct speaking engagements separated in time and context. The earlier measurement (Case Study 002) analyzed approximately 15 paragraphs from Trump's address to the 74th Session of the United Nations General Assembly. The current measurement analyzes the full text of Trump's remarks at a National Republican Congressional Committee dinner in 2026.
The results are striking. The Beverly Index dropped from 66 to 56. The Psychotic sector activation increased tenfold, from 3.0% to 30.2%. The Optimal sector — which measures linguistic markers of executive function, goal-directed reasoning, and adaptive problem-solving — collapsed from 34.1% to 11.3%. The Neurocognitive sector rose from 2.8% to 10.0%. The Dissociative sector rose from 0.5% to 8.5%. The Neurodegenerative sector more than doubled from 3.3% to 8.0%.
NOUS did not know who was speaking. It read the words and produced these measurements.
| Metric | UNGA (CS-002) | NRCC (CS-005) | Delta | Direction |
|---|---|---|---|---|
| Beverly Index | 66 | 56 | -10 | DECLINED |
| CAP Level | HIGH | MODERATE | — | Reclassified |
| GAD % | 25.2% | 26.5% | +1.3 | Stable |
| Coherence | 0.73 | 0.74 | +0.01 | Stable |
| Pattern | STUCK | STUCK | — | Unchanged |
| Integration | 50 | 50 | — | Unchanged |
| Confidence | 85% | 85% | — | Unchanged |
| S2 — Anxiety | 51.2% | 43.5% | -7.7 | Improved |
| S1 — Mood | 15.9% | 29.1% | +13.2 | ELEVATED |
| S3 — Psychotic | 3.0% | 30.2% | +27.2 | CRITICAL SHIFT |
| S4 — Neurocognitive | 2.8% | 10.0% | +7.2 | ELEVATED |
| S6 — Optimal | 34.1% | 11.3% | -22.8 | COLLAPSED |
| S5 — Developmental | 0.8% | 9.9% | +9.1 | ELEVATED |
| S8 — Neurodegenerative | 3.3% | 8.0% | +4.7 | ELEVATED |
| S11 — Dissociative | 0.5% | 8.5% | +8.0 | ELEVATED |
| S12 — Personality | 11.1% | 8.0% | -3.1 | Slightly down |
| S10 — Substance | 12.2% | 5.3% | -6.9 | Improved |
| S7 — TBI/CTE | 5.0% | 5.0% | 0.0 | Unchanged |
| S9 — Captivity | 2.6% | 1.7% | -0.9 | Stable |
All values are unmodified NOUS platform output. Presidents measured March 26, 2026 (CS-001, CS-002). Mass murderers measured March 26, 2026 (CS-003). Trump NRCC measured March 27, 2026 (CS-005). Dashes indicate values not recorded in the original case study.
| Metric | Lincoln | Obama | Trump UNGA | Trump NRCC | Kaczynski | Breivik | Rodger |
|---|---|---|---|---|---|---|---|
| Beverly Idx | 58 | 64 | 66 | 56 | 89 | 100 | 72 |
| CAP | MOD | MOD | HIGH | MOD | CRIT | CRIT | HIGH |
| GAD % | 21.1 | 28.6 | 25.2 | 26.5 | 52.0 | 63.8 | 32.9 |
| Coherence | 0.77 | 0.74 | 0.73 | 0.74 | 0.66 | 0.58 | 0.73 |
| Pattern | STUCK | STUCK | STUCK | STUCK | HYPER | HYPER | STUCK |
| S2 Anxiety | 38.2 | 22.5 | 51.2 | 43.5 | 78.2 | 100 | 57.2 |
| S1 Mood | 20.6 | 23.3 | 15.9 | 29.1 | 61.6 | 66.9 | 24.7 |
| S3 Psychotic | 9.7 | 10.0 | 3.0 | 30.2 | 8.8 | 8.1 | 19.8 |
| S4 Neurocog | 10.0 | 18.0 | 2.8 | 10.0 | — | — | — |
| S6 Optimal | 21.8 | 14.4 | 34.1 | 11.3 | 7.2 | 7.2 | 11.0 |
| S12 Personality | 6.5 | 9.6 | 11.1 | 8.0 | 42.8 | 43.4 | 9.7 |
| S8 Neurodeg | 8.0 | 8.0 | 3.3 | 8.0 | — | — | — |
| S11 Dissoc | 0.5 | 0.5 | 0.5 | 8.5 | — | — | — |
| S5 Develop | 5.4 | 6.9 | 0.8 | 9.9 | — | — | — |
| Condition | Probability |
|---|---|
| GAD (Generalized Anxiety Disorder) | 26.5% |
| Schizoaffective Disorder | 20.1% |
| Bipolar Disorder | 20.0% |
| Psychotic Spectrum Disorder | 18.0% |
| MDD (Major Depressive Disorder) | 17.9% |
| PTSD | 8.8% |
| Parkinson's Disease (Prodromal) | 8.0% |
| VCI (Vascular Cognitive Impairment) | 8.0% |
| Stroke Risk (Cerebrovascular) | 6.8% |
In the UNGA address, Trump's Optimal sector was 34.1% — the highest of any president measured, and a signature of his rhetorical style: direct, declarative, command-oriented language that NOUS reads as executive confidence. In the NRCC remarks, Optimal dropped to 11.3%. That is a 67% reduction.
For context, the mass murderers in Case Study 003 ranged from 7.2% to 11.0%. Trump's NRCC Optimal score now sits inside the pathological range, not because the content is pathological, but because the linguistic architecture of executive function — structured goals, clear directives, organized reasoning — has been replaced by something else entirely: anecdotal rambling, self-referential digressions, and associative topic shifts. This is a structural change in the language, not a content change.
From 3.0% to 30.2%. This is the single largest sector shift in the entire NOUS case study series, including the mass murderer comparisons. To be clear: the Psychotic sector does not measure "psychosis" in the clinical sense. It measures linguistic markers associated with thought disorder: loose associations, tangential reasoning, referential shifts without grounding, and logical non-sequiturs.
The NRCC transcript contains numerous examples: mid-sentence topic abandonment, claims without logical connectives, and rapid oscillation between unrelated subjects (wars, real estate aesthetics, voter ID, midterm history, Iran negotiations, Mike Johnson's family) without transitional structure. NOUS read this correctly. The prior UNGA speech, even with Trump's characteristic style, maintained logical thread continuity. This one does not.
From 15.9% to 29.1%. Trump's UNGA address was remarkably low on Mood activation because it framed challenges as problems to solve through strength, not losses to mourn. The NRCC remarks carry a different emotional signature: nostalgia, in-group warmth, grievance, and affective loading. "We feel so thrilled... then it goes down to one... then we feel rich again" — this language oscillates between emotional highs and lows within a single passage. NOUS detected that the emotional register has widened and become more labile. This is consistent with the Bipolar Disorder diagnostic probability rising to 20.0%.
Neurocognitive rose from 2.8% to 10.0%. Neurodegenerative rose from 3.3% to 8.0%. These sectors measure linguistic markers associated with cognitive processing difficulty and markers associated with progressive neurological decline, respectively.
The NRCC transcript contains features that activate both: word-finding circumlocutions ("For whatever reason, I don't know what it is"), abandoned clauses, repetitive phrasing structures, and reduced syntactic complexity compared to the UNGA address. The diagnostic table now includes Parkinson's Disease (Prodromal) at 8.0%, VCI at 8.0%, and Stroke Risk at 6.8%. These were absent from the UNGA analysis. NOUS is not diagnosing these conditions. It is reporting that the language contains markers these conditions are known to produce.
From 0.5% to 8.5%. The Dissociative sector was at absolute floor in the UNGA speech and across all three presidential measurements. The NRCC speech activated it significantly. Dissociative markers in language include temporal disorientation ("16 months ago... we saved our country"), identity fluidity between self and group ("we feel... they stick together... I've gone to people"), and narrative fragmentation where the speaker shifts referential frame without signaling. The speech moves between first person singular, first person plural, and third person without consistent grounding.
The UNGA address was a formal, teleprompter-delivered speech at the United Nations. The NRCC remarks, while teleprompters were present, reflect Trump's well-documented pattern of departing from prepared text. NOUS has now measured the same speaker in both modes. The difference is not subtle: 10 points on the Beverly Index, a tenfold increase in Psychotic sector activation, and the near-total collapse of Optimal function.
This is precisely what a cognitive signature instrument should detect. When a speaker departs from prepared text and produces spontaneous language, the cognitive architecture of that language changes. NOUS measured the change. The instrument is sensitive to mode of production, not just content — which is a core claim of Neural Oscillation Signature Theory.
Trump's NRCC Beverly Index of 56 now sits below all three presidential baselines (Lincoln 58, Obama 64, Trump UNGA 66). His Optimal score of 11.3% is now closer to the mass murderer range (7.2–11.0%) than to his own prior measurement (34.1%). His Psychotic activation of 30.2% exceeds every subject in the series except the three mass murderers' average.
However, and this is critical: the Beverly Index did not drop into the pathological range. It is 56, still MODERATE. The Coherence score actually improved slightly (0.74 vs 0.73). The Pattern remains STUCK, not HYPER. NOUS is not producing a pathological reading. It is producing a reading of cognitive deterioration within the normal range — exactly the kind of subtle, longitudinal signal that the platform was designed to detect.
Case Study 005 is the first longitudinal measurement in the NOUS case study series. It demonstrates something no cross-sectional comparison can: that NOUS detects change within the same individual over time. The platform measured Trump at the United Nations and produced one cognitive signature. It measured Trump at a party fundraiser and produced a measurably different one. The differences are not random. They map precisely to what linguists, neuropsychologists, and political analysts would predict: less structure, more associative drift, reduced executive markers, increased emotional lability, and the emergence of markers associated with cognitive processing difficulty.
This is the VOCA pathway proof of concept. VOCA — Voice-Over-Time Cognitive Analysis — is designed to track cognitive signatures longitudinally. Case Study 005 is the first empirical demonstration that the platform can do exactly that.
The subject did not consent, was not in a clinical setting, and provided no metadata. NOUS read the words and measured the change.
For evaluators reviewing this for the ONI solicitation (N6133126SNQ18), the DARPA CLARA proposal (DARPA-PA-25-07-02), or the Frontiers peer review (Manuscript ID 1834364): this case study shows that NOUS produces consistent, interpretable, and longitudinally sensitive measurements from publicly available text, without subject cooperation, without metadata, and without human scoring bias.
The instrument works. And it works on one of the most analyzed speakers in the English language.
The cognitive signatures, NOSS profiles, and Beverly Index scores in this study were generated by the NOUS cognitive intelligence platform. Licensed practitioners can access NOUS directly for clinical cognitive assessment.