We processed 1,,,, words of running text and are publishing the counts for all 1,,, five-word sequences that appear at least 40 times.
There are 13,, unique words, after discarding words that appear less than times. And let us hear from you - we're excited to hear what you will do with the data, and we're always interested in feedback about this dataset, or other potential datasets that might be useful for the research community. Furthermore, taking unequal weights and time trends simultaneously into account, the significant positive trend remains only for our set of original terms see Table 3 , Model VII but turns negative for our set of synonyms.
Finally, combining words of comparatively high frequency and the most advanced standardization procedure with an additional set of synonyms, no significant trend is observable any longer see Table 4 , Model VII.
Finally, by the example of forced secularization during the Soviet regime, we draw attention to the danger of biased estimations that can arise from censorship and propaganda. However, we would like to point out that what we believe is censorship might be also attributable to the lack of metadata. In particular, Koplenig [ 44 ] shows that with the lack of proper metadata, it cannot be ruled out that trends arise due to changes in the composition of the underlying data.
Thus, some of the trends we observe for the Russian corpus may not necessarily result from censorship as part of forced secularization, but some change in the data. By applying all of the above-mentioned procedures to religious 1-grams, this study further contributes to the body of research that investigates the development of collectivism and individualism see, e. In particular, we study the cross-cultural development of religious trends for the years to , with a particular focus on the development in times of crisis such as WWII.
Except for the Italian corpus, our analyses reveal a relative overall decrease for religious terms. The trend towards an increased expression of religion during a severe time of crisis is, however, not only observable in the German corpus.
Additional analyses further reveal a significant positive and robust trend for Italian religious terms. In contrast, for the American and British English corpora we do not find such a reversal but a constant and significant negative time trend over the whole observation period.
Our study is not without limitations. Although the developers have addressed this issue by providing a corpus with a better-balanced text collection, i. Second, we used the new updated corpora to exploit the advantages of improved OCR and better underlying library and publisher metadata. However, as suggested by Twenge et al. We therefore re-examined all analyses using the old corpora.
Overall, previous findings are confirmed. However, there was no Italian corpus before the update. Finally, we agree that the key assumption of Google Ngram studies, i. Overall, Google Ngram has allowed scholars to shed further light on various topics such as gender differences [ 17 , 18 , 19 ], emotions [ 20 , 21 , 22 , 45 ], personality [ 23 , 24 ], cognition [ 25 , 26 , 27 ], hypnosis and psychotherapy [ 46 ], moral values [ 47 ], education [ 48 ], nature [ 49 ], astrology and phrenology [ 50 ], and the development of individualism and collectivism e.
In this respect, despite limitations, we believe that Google Ngram is a beneficial tool for research purposes and that the procedures presented in this guideline can reinforce the reliability of derived results. Browse Subject Areas? Click through the PLOS taxonomy to find articles in your field. Abstract The Google Books Ngram Viewer Google Ngram is a search engine that charts word frequencies from a large corpus of books and thereby allows for the examination of cultural change as it is reflected in books.
Introduction Since its launch in , the possibilities and limitations of using the Google Books Ngram Viewer Google Ngram for research purposes have been controversially discussed. Assessing the development of religion via Google Ngram While the world population almost quintupled from approximately 1. Method and results This section describes how to conduct a Google Ngram study.
Collection of words and verification procedure We obtain a set of 23 religious English terms from Ritter and Preston [ 35 ] who surveyed the literature of words that were used to prime religious concepts. Download: PPT. Table 1. List of religious English terms with their German and Italian translations.
Baseline analysis After extracting word frequencies from Google Ngram, prior studies have investigated cultural changes by examining the correlation coefficients between word frequencies and years. Procedure I: Multiple languages As highlighted by our current analysis, one big advantage of Google Ngram is the possibility to compare cultural changes in a cross-cultural setting. Fig 1. Table 2. Overview of original words and their higher frequency inflections.
Procedure IV: Synonyms To verify that investigated words reflect true underlying concepts rather than idiosyncrasies, re-checking initial findings with several synonyms is a strong robustness check.
Procedure V: Standardization of word frequencies In spite of Michel et al. Table 3. Correlation coefficients for different standardization procedures. Composite analysis In this section, we combine our suggested procedures by using higher frequency words, synonyms, and the standardization procedure that accounts for unequal weights and the influx of data.
Table 4. Correlation coefficients for different standardization procedures using higher frequency words and synonyms. Censorship and propaganda Finally, we discuss the importance of taking any potential censorship and propaganda into account. Discussion Google Ngram allows for hands-on quantification of cultural change using millions of books.
Supporting information. S1 Appendix. References 1. Gooding P. Mass digitization and the garbage dump: The conflicting needs of quantitative and qualitative methods. Literary and Linguistic Computing , 28 3 , — View Article Google Scholar 2. Pechenick E. Characterizing the Google Books corpus: Strong limits to inferences of socio-cultural and linguistic evolution. PloS One , 10 10 , e Pettit M. Historical time in the age of big data: Cultural psychology, historical change, and the Google Books Ngram Viewer.
History of Psychology , 19 2 , Michel J. Quantitative analysis of culture using millions of digitized books [With Supporting Online Material]. Science , , — Lin Y.
Syntactic annotations for the Google Books Ngram corpus. Association for Computational Linguistics. Twenge J. Increases in individualistic words and phrases in American books, — PloS One , 7 7 , e View Article Google Scholar 7. Changes in pronoun use in American books and the rise of individualism, — Journal of Cross-Cultural Psychology , 44 3 , — View Article Google Scholar 8. Kesebir P. The cultural salience of moral character and virtue declined in twentieth century America.
The Journal of Positive Psychology , 7 6 , — View Article Google Scholar 9. The seven words you can never say on television: Increases in the use of swear words in American books, — SAGE Open , 7 3 , View Article Google Scholar Greenfield P.
The changing psychology of culture from through Psychological Science , 24 9 , — Hamamura T. Changes in Chinese culture as examined through changes in personal pronoun usage. Journal of Cross-Cultural Psychology , 46 7 , — Folk beliefs of cultural changes in China. The enablements of the Google Books Ngram Viewer provide complementary information sourcing for designed research questions as well as free-form discovery.
The tool also enables the download of raw dataset information of the respective ngrams, and the findings are released under a generous intellectual property policy. This presentation will introduce this semi-controversial tool and some of its creative applications in research and learning. Who are you? Any research angles that may be informed by n-grams? What is your level of experience with this Viewer? What is the Google Books Ngram Viewer?
History cont. Some Terminology cont. Anomalous Years of Interest; Links 15 Downloadable Experimental Datasets 17 Research Potential? What sorts of professions lead to fame?
What is the trajectory of collective consciousness from knowing to not knowing? How do cultural phenomena affect human populations over time? How do rules of language become normalized? Two Element Comparison cont. Chicken or the Egg? Once the basic extractions are acquired, try the more complex ones using tagging and combinatorial approaches. TMX — this format is only availble with parallel copora. Preloaded corpora Preloaded corpora in Sketch Engine cannot be downloaded but word embeddings computed from these corpora for the purpose of language modelling and similar applications are available for download from our word embeddings page.
On-demand corpus building request language data or service. Please enable cookie consent messages in backend to use this feature.
0コメント