1 / 28
Navigate Esc Close
Scroll
NLP Analysis of Political Discourse

Macron and French Diplomacy

What the ambassadors speech analysis reveals

January 8, 2026

Antoine Lemor | NLP-POL PROJECT ↗

About the project

What is NLP-POL?

NLP-POL is a project of recurring analyses in political science, applying Natural Language Processing (NLP) methods to political and policy issues.

Objective

Demonstrate the possibilities offered by NLP and AI annotation methods. Democratize these approaches.

Tools

Transcribe-tool ↗, LLM_Tool ↗, nlp-pol ↗

Frequency

Regular publications on political affairs, shared on LinkedIn and my blog ↗

Methodology

Annual Ambassadors Conference at the Elysee - January 8, 2026

📝

Transcription

with

Transcribe-tool

Tokenization

with

Transcribe-tool
🤖

LLM Annotation

with

LLM_Tool
📊

Analysis

with

nlp-pol
359
Sentences analyzed
119
Political sentences
560
Actor mentions
264
Geopolitical frames
Dashboard

Finding: a doctrine of pragmatic assertion

Overview

Balanced vision

A GAI of +0.08 reveals a vision that is neither alarmist nor euphoric. Macron frames challenges without catastrophism.

+0.08 Geopolitical index

France as actor

With 85% agency, France is presented as proactive, not reactive to events.

85% Agency

Programmatic policies

45% concreteness: proposals remain at strategic rather than operational level.

45% Concreteness

Neutral tone

A tonality close to zero (-0.01) indicates a pragmatic register, neither alarmist nor triumphalist.

-0.01 Tonality
Analysis #1

The speech at a glance

Emotional timeline

Finding: dominant pragmatism, targeted combative peaks

Emotional register evolution

Dominant register

Pragmatic (30%) and confident (21%) tones dominate the entire speech.

51% Pragmatic + Confident

Combative peaks

12% of the speech adopts a combative tone, focused on specific subjects (defense, sovereignty).

12% Combative

Alarmist passages

Only 7% of the speech is alarmist - reserved for the most serious threats.

7% Alarmist

Overall tonality

The score of -0.01 indicates a balance between positive and negative registers.

-0.01 Tonality index
Analysis #2

What vision of the world?

Geopolitical vision

Finding: opportunities before threats

Geopolitical framing

Threat frames (39%)

  • Power politics - 41 occurrences (16%)
  • Global disorder - 17 occurrences (7%)
  • Multilateral decline - 12 occurrences (5%)
  • Fragmentation - 11 occurrences (4%)
  • Vassalization - 7 occurrences (3%)

Opportunity frames (61%)

  • Cooperation - 58 occurrences (22%)
  • Resilience - 30 occurrences (12%)
  • Progress - 15 occurrences (6%)
  • Leadership opportunity - 14 occurrences (5%)
  • Multilateral renewal - 10 occurrences (4%)
Interpretation
The "cooperation" frame dominates: Macron presents challenges as partnership opportunities rather than existential threats.
Analysis #3

How are actors presented?

Actor sentiment

Finding: allies valued, major powers nuanced

560 mentions analyzed
Actor Mentions Net sentiment Positive Negative
France166+0.671154
Europe/EU61+0.51343
United States17-0.0645
China12-0.4205
Ukraine8+0.7560
India8+0.7560
Interpretation
Neutral treatment of the USA (-0.06), negative of China (-0.42). Ukraine and India emerge as privileged strategic partners.
Analysis #4

What policies are proposed?

Policy matrix

Finding: diplomatic method first

119 sentences classified with political content

Specificity level

% of sentences with political content

  • Programmatic - 67 sentences (56%)
  • Aspirational - 48 sentences (40%)
  • Concrete - 4 sentences (3%)

Very few concrete and operational measures.

Priority domains

% of sentences with political content

  • Diplomatic method - 29 sentences (24%)
  • European economy - 9 sentences (8%)
  • Global governance - 8 sentences (7%)
  • Africa relations - 7 sentences (6%)
  • Trade policy - 7 sentences (6%)
Interpretation
The emphasis on "diplomatic method" (24% of political sentences) reveals a priority on how rather than what: the way of negotiating takes precedence over content.
Analysis #5

What rhetorical strategy?

Rhetorical strategy

Finding: balance between diagnosis and action

537 speech acts

Speech acts

  • Stating - 109 (20%)
  • Proposing - 76 (14%)
  • Exhorting - 71 (13%)
  • Framing - 69 (13%)
  • Reassuring - 67 (12%)

Action orientation

Action / (action + diagnostic) ratio:

43%

Balance between situation analysis and calls to action

Interpretation
The speech combines diagnosis (stating, framing) and voluntarism (proposing, exhorting) in a near 50/50 balance.
Analysis #6

How is France positioned?

Agency profile

Finding: active agent, not victim

304 positionings
193
Active
78
Cooperative
33
Reactive

Active agent (46%)

France is presented as initiating actions, not reacting to events.

Partner (23%)

Emphasis on cooperation and alliances rather than unilateralism.

Leader (8%)

Leadership positioning on key issues (Europe, climate, defense).

Victim (only 3%)

Near-absence of victimhood positioning - no complaint or fatalism.

Analysis #7

What diplomatic positioning?

Diplomatic positioning

Finding: Europe first, major powers nuanced

Actor groups

Allies (France, Europe)

Systematically positive treatment. European anchoring affirmed.

+0.59Average sentiment

Major powers

USA slightly negative (-0.06), China negative (-0.42). No automatic alignment.

-0.24Average sentiment

Emerging partners

Ukraine and India as new privileged strategic partners.

+0.75Average sentiment

Notable absence

Russia is rarely mentioned directly - framing by consequences rather than confrontation.

Conclusion

A doctrine of pragmatic assertion

Quantitative analysis reveals a diplomacy that combines realism about threats and voluntarism in action. Macron positions France as a proactive actor in a changing world, favoring partnerships and multilateral cooperation.

+0.08
Cautious vision
85%
Agency
61%
Opportunity frames

Antoine Lemor

github.com/antoinelemor ↗

Open Source Tools

📊

NLP Analysis

nlp-pol
📝

Transcription

Transcribe-tool
🤖

LLM Annotation

LLM_Tool