Speeches at the United Nations General Assembly often reach for moral authority by citing historical figures. We analyzed speech "chunks" (text segments) from 1950 to 2024 to identify who is most often quoted.
Top 20 Overall (1946-2024)
View all 179 figuresThe 2010s: Legacy & Icons
View Analysis →A decade defined by retrospection and tributes. The passing of Kofi Annan and Nelson Mandela turned their words into guiding principles for the assembly.
The 2020s: Crisis & Urgency
View Analysis →Facing a pandemic and global instability, speakers shifted to voices of urgency. Churchill's crisis leadership and calls for racial justice invoking MLK.
The 2000s: Millennium & Hope
View Analysis →The turn of the millennium brought a focus on new beginnings. Kofi Annan's influence was paramount, while Gandhi and King remained touchstones.
The 1990s: New World Order
View Analysis →The fall of the Berlin Wall and the end of the Cold War ushered in unprecedented optimism. Nelson Mandela's release and the Oslo Accords defined a decade of hope.
The 1980s: Cold War Twilight
View Analysis →The final decade of the Cold War saw superpower tensions and the rise of Gorbachev's reforms. Nelson Mandela became a global symbol of resistance.
The 1970s: Detente & Upheaval
View Analysis →A decade marked by superpower detente, the oil crisis, and the end of the Vietnam War. China's entry into the UN in 1971 reshaped global diplomacy.
The 1960s: Decolonization & Crisis
View Analysis →The decade of African independence saw dozens of new nations join the UN. The Cuban Missile Crisis brought the world to the brink.
The 1950s: Early Cold War
View Analysis →The formative decade of the United Nations, shaped by the Korean War and the dawn of the atomic age. Dag Hammarskjöld became the iconic Secretary-General.
Research Methodology
1. Building the Figure Database
We curated a database of 179 notable figures across 10 categories: philosophers, political thinkers, economists, scientists, writers, historical figures, civil rights leaders, religious figures, UN leaders, and world leaders. For each figure, we store name variations to catch different spellings and references (e.g., "Mahatma Gandhi", "Mohandas Gandhi", "M.K. Gandhi").
2. Extracting Mentions
We analyzed over 120,000 text segments ("chunks") from UN General Assembly speeches spanning 1946-2024. Each chunk was searched for mentions of our notable figures using pattern matching against their name variations.
3. Identifying Direct Quotations
Not every mention is a quote. To distinguish actual quotations from mere references, we look for attribution patterns where the quoted text is explicitly attributed to the figure. This prevents false positives like marking random quoted text near a name mention as a "quote by" that person.
# Patterns for quotes ATTRIBUTED to a figure
# "Gandhi said '...'"
r'{name}\s+said[,:]?\s*["'](.+)["']'
# "'...' - Gandhi"
r'["'](.+)["']\s*[-–]\s*{name}'
# "As Gandhi wrote, '...'"
r'as\s+{name}\s+wrote[,:]?\s*["'](.+)["']'4. Confidence Scoring
Each mention receives a confidence score from 0.3 to 1.0:
- 0.95 — Clear attribution with quoted text (e.g., "Gandhi said, 'Be the change...'")
- 0.5 — Weaker indicators (e.g., "according to Gandhi", "Gandhi believed")
- 0.3 — Simple mention without attribution context
Only mentions with clear attribution patterns (confidence ≥ 0.9) are marked as "direct quotes" and displayed in the quotation sections. All other mentions are tracked but shown separately.
5. Full Transparency
Every quotation links directly to its source speech. Click on any figure's name to see all their quotations with the exact speech, year, and country where each quote was cited. This allows you to verify any data point against the original source material.
CREATE TABLE notable_figures (
id INTEGER PRIMARY KEY,
name TEXT NOT NULL,
category TEXT NOT NULL
);
CREATE TABLE quotations (
id INTEGER PRIMARY KEY,
figure_id INTEGER,
speech_id INTEGER,
quote_text TEXT,
confidence REAL
);Limitations
This methodology has some inherent limitations:
- Paraphrased quotes without quotation marks are not captured as direct quotes
- Misattributed quotes (where a speaker incorrectly attributes a quote) are included as-is
- Some figures with common names may have fewer results due to stricter matching to avoid false positives
- Translations of speeches may vary in how they render quotations