Journal of

Journal of Language Research

Current Issue Volume No: 1 Issue No: 1

ISSN: 2998-4122
share this page

Research Article Open Access
  • Available online freely Peer Reviewed
  • Provisional

    A stylometric analysis of two sentimental letters by Albert Einstein and Simón Bolívar

    Raul Isea 1  

    1Fundación IDEA Hoyo de la Puerta, Baruta, Venezuela.


    On social networks often circulate emotional letters from various authors that appear after their death. The supposed letter Albert Einstein wrote to his daughter Lieserl and the letter Simon Bolivar wrote to his cousin Fanny are two examples. A Python-developed tool is used to do a stylometric study to determine whether these authors were. The language styles of Einstein and Bolivar were described using eight and six letters, respectively. The results show that they never wrote these letters.

    Author Contributions
    Received 04 Sep 2023; Accepted 30 Oct 2023; Published 29 Nov 2023;

    Copyright © 2023 Raul Isea

    Creative Commons License     This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

    Competing interests

    The authors have declared that no competing interests exist.


    Raul Isea (2023) A stylometric analysis of two sentimental letters by Albert Einstein and Simón Bolívar. Journal of Language Research - 1(1):15-20.

    Download as RIS, BibTeX, Text (Include abstract )

    DOI 10.14302/issn.2998-4122.jlr-23-4739


    Letters written by great people in history are circulating on social networks. An example is the letter that the German physicist Albert Einstein (1879-1955) wrote to his daughter Lieserl Einstein. In this letter, he shows that love is the only answer to the survival of our species. There is also a letter that Simon Bolívar (1783-1830) claims to have written to his cousin Fanny du Villars, written on her deathbed, telling her how important she was in his life.

    In fact, Albert Einstein's letter to his daughter analyzes the concept of love as a universal force that science has not yet fully developed. This letter says that love is seen as light, gravity and life. It is the only energy in the universe that humans have not learned to control at will. Suggests that love is the most powerful force because it has no limits.

    Simón Bolívar wrote a letter to his cousin Fanny du Villars, obtained from the archives of the National Academy of Medicine of Colombia (, to express his emotional affection for her. In the letter, he describes the surrounding nature and emphasizes the beauty of the Caribbean Sea, the magnificent mountains and the colorful sky, and tells her how important she was in his life. The strange thing is that the real name of the French woman was Louise Jeanne Nicole Arnalde "Fanny" Denis de Keredern de Trobriand (1775-1859).

    On the other hand, all the papers of Albert Einstein are collected in various databases such as The Collect Papers of Albert Einstein (, Einstein Papers Project (, and so on. None of them have any proof of the existence of this letter to his daughter.

    The same applies to the documents of Simón Bolívar available at Biblioteca Virtual Miguel de Cervantes ( There is also no information in these files about the existence of this emotional letter. For all this, these letters are analyzed according to the stylometric methodology as described below.


    Stylometric has been summarized as a statistical methodology based on the frequency analysis of each author's own words 14, 6, 11. The term was coined by the Polish writer Wincenty 8 when he determined the chronology of Plato's dialogues 8 and was originally based on a chi-square analysis of word length-word and their frequency 9.

    More recently, it has been used to determine the authorship of some unknown or questionable works, such as a comedy originally attributed to Miguel Bermúdez by the National Library of Spain, but found by this method to be the author of Lope de Vega. 3.

    Another example was Ulysses, published in 1922 by the Irishman James Joyce, but stylometric studies conclude that it was written by five other people who were not mentioned in the work. 12.

    Based on the above, the paper analyzes eight letters written by Albert Einstein and finds out if the letter to his daughter was really written by him. This procedure is repeated in a letter that Simón Bolívar allegedly wrote to his cousin, as described below.


    Daelemans 4 describes the methodology for performing a stylometric analysis and is summarized below. A corpus is created from the letters of the same author. The selected letters and the numbers were chosen at random (details in Table 1).

    Table 1. Details of the letters written in English by Albert Einstein, and the letters in Spanish by Simón Bolívar.
    Abbreviation Who wrote the letter Date To whom the letter is addressed
    B362 Bolívar 1830, May 26 General Sucre
    B363 Bolívar 1830, May 26 Juan de Dios Amador
    B364 Bolívar 1830, May 31 Juan de Dios Amador
    B365 Bolívar 1830, Jun 17 Pedro Medrano
    B367 Bolívar 1830, Jul 31 Manuela Garaycoa de Calderón
    B371 Bolívar 1830, Oct 17 Joaquín de Mier
    Fanny Bolívar 1807, Sep 14 Fanny du Villers
    Freud Einstein 1932, Jul 30 Sigmund Freud
    God Einstein 1954, Jan 3 Mr. Gutkind
    Borns Einstein 1924, Apr 29 Max Born
    Roosevelt Einstein 1945, Mar 25 F. D. Roosevelt
    Curie Einstein 1911, Nov 23 Marie Curie
    Switzer Einstein 1953, Apr 23 J. S. Switzer
    Palestine Einstein 1948, Apr 10 Shepard Rifkin
    Szilard Einstein 1939, Aug 2 F. D. Roosevelt
    Daughter Einstein ¿1903,Sep 19?* Lieserl Einstein

    (*) There is no agreement at this dated.

    For each author, the word frequency is determined according to the Delta function, which is a linguistic measure capable of distinguishing the authorship of texts according to the definition proposed by 2.

    Some articles questioned the use of the Delta function, but Burrows showed that it is an excellent technique for identifying the authorship 13. This function simply determines the frequency variation of the most frequent words in the text using z-scores 13. Remember that z-scores are a measure of relative frequency difference minus the word mean divided by the standard deviation 16.

    After that, a matrix was calculated based on the distance obtained from the frequency of the words 15. This calculation uses one, two or more words, and this number is indicated as n-grams 15.

    Based on this frequency, the distance Remember that there were two possible element types used to generate n-grams: words and characters. Character n-grams show how frequently certain letters, capital letters, punctuation marks, or numbers are used at the alphabetic level of a language, while the word n-grams and vocabulary in a document are connected. In addition to word frequency, these characteristics also include sentence length, word length distribution, richness of vocabulary, and lexical mistakes. These can be used as the initial tokenization step for any language 10.

    For other part, it is usually calculated using the Manhattan, Euclidean, and so on 15. The result is usually visualized with a dendogram, a tree where similar distances are grouped based on a certain number of words into a single conglomerate or cluster, simply abbreviated as MRW (ie., Meaning Most Words).

    The paper considers that sentences written by the same person should be grouped into the same branch or node 5. All calculations were performed in the Python programming language.


    Figure 1 shows the result of normalization of the frequency of words obtained from the English letters of Albert Einstein when the n-gram is equal to two (ie two words). It can be seen that the alleged letter to daughter does not follow the same pattern as her other letters, with the words with the biggest difference such as “of the”, “it is”, “in a”, “is the”, “and the”, “to me”, among others.

    Figure 1. Normalization of the frequency of appearance of words obtained from Albert Einstein's letters, using the Manhattan distance, and MRW equal to 35.
      Normalization of the frequency of appearance of words obtained from Albert Einstein's letters, using the Manhattan distance, and MRW equal to 35.

    Figure 2 shows the result of the normalization of the frequencies of words obtained from the letters of Simón Bolívar using n-gram equal to 1 and MRW = 50. This graph shows differences between the letters, such as “mi”, “los”, “para”, ”he”, “si”, “dinero”, among others.

    Figure 2. Normalization of the frequency of appearance of words obtained in the letters of Simón Bolívar using Manhattan’s distance.
      Normalization of the frequency of appearance of words obtained in the letters of Simón Bolívar using Manhattan’s distance.

    Figure 3 shows a rectangular dendogram of Albert Einstein's letters. It was observed that the supposed letter to the daughter does not reproduce the linguistic styles of the other eight letters, that is, the supposed letter (labeled Daughter) does not belong to any of the three groups (clusters) that make up the rest of letters.

    Figure 3. Rectangular dendrogram obtained from a stylometric analysis of Albert Einstein's letters, visualized with iTOL (Letunic, 2021).
      Rectangular dendrogram obtained from a stylometric analysis of Albert Einstein's letters, visualized with iTOL (Letunic, 2021).

    Figure 4 shows a rectangular dendogram of Simón Bolívar's letters, showing that his cousin Fanny's letter does not belong to any of the two large conglomerates that repeat the style of Bolivar. Therefore, Fanny's letter was not written by Simón Bolívar.

    Figure 4. Rectangular dendrogram obtained from the stylometric analysis of the Bolívar letters, visualized with iTOL (Letunic, 2021).
      Rectangular dendrogram obtained from the stylometric analysis of the Bolívar letters, visualized with iTOL (Letunic, 2021).


    The paper's goal was to determine whether Albert Einstein and Simon Bolivar actually penned amorous letters to their daughter and cousins, respectively, by looking for linguistic evidence in their letters. It is established that they did not write these tearful letters based on the findings of a stylometric examination.


    1.Bensalem I, Rosso P y Chikhi, S. (2019) On the use of character n-grams as the only intrinsic evidence of plagiarism,Language Resources and Evaluation. 53(3), 363-396.
    2.J F Burrows. (2002) Delta: A Measure of Stylistic Difference and a Guide to Likely Authorship.LiteraryandLinguisticComputing. 17(3), 267-287.
    3.Bvmc. (2014) 23) "Descubierta una comedia inédita de Lope de Vega" Available at
    4.Daelemans W. (2013) Explanation in Computational Stylometry. In:. Gelbukh, A. (eds) Computational Linguistics and Intelligent Text Processing. CICLing 2013. Lecture Notes in Computer Science , Berlin, Heidelberg 7817, 10-1007.
    5.Eder M. (2017) Visualization in stylometry: Cluster analysis using networks,Digital. Scholarship in the Humanities 32(1), 50-64.
    6.Fuller S y O'Sullivan, J. (2017) Structure over Style: Collaborative Authorship and the Revival of Literary Capitalism.Digital Humanities Quarterly. 11(1), 10-17613.
    7.Letunic I y Bork, P. (2021) Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation.Nucleic Acids Research. 49, 293-296.
    8.Lutoslawski W. (1898) Principes de stylométrie appliqués à la chronologie des œuvres de Platon.RevuedesÉtudesGrecques. 11(41), 61-81.
    9.Mendenhall T. (1887) The characteristic curves of composition.Science. 214-237.
    10.Ríos-Toledo G, JPF Posadas-Durán, Sidorov G, Castro-Sánchez N A. (2022) Detection of changes in literary writing style using N-grams as style markers and supervised machine learning. PLoS One. , Jul 17(7), 10-1371.
    11.G B Schaalje. (2011) and others (2011). Extended nearest shrunken centroid classification: A new method for open-set authorship attribution of texts of varying sizes,Literary and LinguisticComputing. 26, 71-88.
    12.Schoenbaum S. (2018) Internal evidence and Elizabethan dramatic authorship; an essay in literary history and method. Octubre 2018) ISBN 196.
    13.Škorić M, Stanković R, N M Ikonić, Byszuk J, Eder M. (2022) . Parallel Stylometric Document Embeddings with Deep Learning Based Language Models in Literary Authorship Attribution.Mathematics 10(5), 838-10.
    14.Stamatatos E. (2009) A survey of modern authorship attribution methods.JASIST. 60(3), 538-556.
    15.Stanikunas D, Mandravickaite J. (2017) y Krilavicius (2017). Comparison of distance and similarity measures for stylometric analysis of Lithuanian texts. CEUR Workshop proceedings [electronic resource]: ICYRIME , Kaunas, Lithuania Vol, 1852.
    16.Stefan E. (2017) Understanding and explaining Delta measures for authorship attribution,Digital Scholarship. in the Humanities, 32(suppl_2), ii4–ii16. .