Speech analysis

from Wikipedia, the free encyclopedia

On the one hand, language analysis is a technical term from linguistics , where it describes the content ( semantics ) and formal ( syntactics ) examination of a text according to the character of the language means used. On the other hand, "language analysis" is used in different senses in philosophy, for example very generally for the analysis of natural language as a sub-project alongside other areas of responsibility or more specifically for research programs that provide for philosophy to be carried out as an analysis of everyday language.

Speech analysis is also used in psychology primarily for personality analysis.

Linguistic analysis in linguistics

The aim of linguistic language analysis is to provide a better insight into the intentions of a text. The text to be analyzed can be a poem, a short story or another literary text, but also a speech, a letter or an entirely different type of text . Speech analysis is a major part of text or corpus analysis . According to the communication model , the type of text and the use of language are determined by the intention , consequently the language analysis must show the dependence of the linguistic and stylistic means on the respective intention of the text.

In linguistics, a distinction is made between several levels of language analysis, which are differently important for a sufficient analysis depending on the type of text.

Semantic analysis

The focus is on the individual parts of speech , i.e. H. the so-called content words are examined with regard to their frequency in the examined text and their semantic content. Essentially, the following is determined:

  • The most prominent and commonly used part of speech and its performance and function,
  • Origin of the vocabulary (e.g. from the ethical, political, craft, everyday area),
  • Type of nouns ( concrete , abstract , foreign word, buzzword, catchphrase, technical terms ),
  • Type of adjectives (descriptive, evaluative),
  • Type of verbs (action, description, modal and auxiliary verbs, passive and active verbs),
  • Expressive value of the words ( denotation , connotation , transferred meaning ).

Syntactic Analysis

The syntactic analysis shows the structure of the sentences and the types of sentences.

  • Types of sentences: statements, questions, commands
  • Set organization: Parataxe , hypotaxis , hybrid

Style analysis

The style analysis determines the rhetorical or stylistic means that are used.

The latter differentiation should be illustrated using the example of the term “to die”: “Eternal peace found” (poetic), “faded” (elevated), “died” (neutral), “made the mosquito” (sloppy), “scraped off” (vulgar ).

Language Analysis in Philosophy

Language analysis as an analysis of colloquial language

The term "language analysis" can be used in different senses in philosophical contexts. Often times he refers to research programs that suggest that the analysis of common everyday language should be the basic method of philosophy. Such theses are particularly typical for the so-called philosophy of normal language - a temporarily important branch of the so-called analytical philosophy , as well as for many representatives of the so-called linguistic turn . Occasionally one also spoke of "language analytical philosophy". The identification of analytical philosophy and language analysis, which is occasionally found in older literature, is, however, wrong in the history of philosophy. The philosophical investigations by Ludwig Wittgenstein , which discuss various linguistic practices, not just the formulation of statements, became particularly important for this normal language philosophy .

Language analysis as an orientation towards formal languages

Distinguishable from the project of normal language philosophy is the concern, which is also widespread in parts of analytical philosophy, to base the philosophical discussion and the reconstruction of the content of linguistic utterances on formal languages (so-called philosophy of ideal language ).

History of Speech Analysis

An analysis of everyday language and also of the special forms of expression used among technical philosophers is already carried out by philosophers of ancient Greece, for example by Socrates and Plato . A reconstruction of the content and the references or truth makers of statements by means of the linguistic-philosophical and logical terms of Aristotle is particularly formative in terms of the history of the impact . A single topic of the ancient “language analysis” is the question under which conditions which types of linguistic utterances can be judged as true or false, the so-called “truthfulness” especially of statements.

Language analysis in psychology and human resources

The aim is to use the linguistic peculiarities of a text to affect personality , mood or mental illnesses such as B. to conclude a depression of the author. James Pennebaker is a pioneer in this area . There are providers who create a personality profile through a language analysis and use this as a basis for personnel selection or for personnel development. From a scientific point of view, these procedures are currently to be classified as dubious, as their claims are empirically not solidly proven. The connection between language / voice and personality is minimal and the connection with professional success has not been proven. This means that such instruments are not serious instruments at least at the moment.

Nevertheless, there are software products on the market that make such promises. Precire Technologies GmbH in Aachen received the negative BigBrotherAward in 2019 "for its scientifically dubious, probably illegal and dangerous language analysis". Precire is not only used to pre-select female applicants, but also to analyze the emotions of people who call a hotline.

Speech analysis in customer communication

Speech analysis enables the categorical evaluation of recorded telephone calls between a company and its customers. It offers extended functionality and valuable information from customer calls. This information can be used to gain insights into strategy, product, process and operational issues as well as the performance of the contact center agents. Speech analysis can also identify areas where contact center agents need additional training or coaching, and automatically monitor customer service provided in calls. Speech analysis providers use processing modules from third parties or develop their own processing modules.

The process of language analysis can isolate the most frequently used words and phrases over a period of time and determine whether usage is trending up or down. This information is useful for supervisors, analysts and other functionaries in a company to identify changes in customer behavior, initiate measures to reduce the call volume and increase customer satisfaction. They provide insight into the customer's thought processes and thus enable companies to make corrections.

Call centers are the interfaces for customer communication. Speech analysis has a purpose of its own here. It is less about stylistic analyzes, cultural developments or philosophical interpretations of what has been said. Customer communication, i.e. professional dialogue, is about taboo phrases, reference phrases, catchphrases, softeners, etc. - that is, vocabulary that can have a positive or negative influence on a conversation with a customer. The language in a call center is analyzed to determine whether and which such terms were used in conversations by whom and in what context. One possible way is e.g. B. the recording of the conversations, the subsequent writing of the voice files.

Examples of the metrics of speech analysis in customer communication

  • Taboo phrases: Either the rules of professional dialogue or the specifications of a project exclude the use of certain words or phrases. Typical taboo phrases are therefore “I'm not responsible for that”, “You would have to please first” or “I can't help you”. Names of competing companies or insults etc. could also be defined as taboo phrases. Here, a voice analysis in the call center takes care of the tracking and the coaching of the agents that is made possible to improve dialogue skills.
  • Target phrases: friendliness and service orientation are measurable. In the speech analysis in call centers, phrases like “very much,” “I'll take care of” or “thank you” are tracked in order to find out which agents represent this attitude and which do not. Here, too, it is again possible to check project-related specifications.
  • Softener: A certain amount of commitment is important, especially in sales telephony. Words like “actually”, “maybe”, “possibly” or generally subjunctive should be avoided. These are also found with the help of the speech analysis of call recordings and can be used together with the appropriate voice files for training the agents.

literature

  • Ernst Tugendhat : Lectures as an introduction to the philosophy of language analysis. Suhrkamp, ​​Frankfurt a. M. 1976, ISBN 3-518-27645-X (Engl. Traditional and Analytical Philosophy: Lectures on the Philosophy of Language. Ins Engl. Translated by PA Gorner, CUP, Cambridge, Mass. 2010, ISBN 978-0-521 -12573-4 )

Web links

Wiktionary: Language analysis  - explanations of meanings, word origins, synonyms, translations

Individual evidence

  1. ↑ You are what you speak on Spektrum.de, accessed on June 20, 2015
  2. ↑ Winfried Löffler, for example, discusses some reasons for this error, which can occasionally still be found in older or non-specialist literature: Who is afraid of analytical philosophy? A relationship that is still clouded, in: Voices of Time 6 (2007), pp. 375–388.
  3. ↑ This mentions e.g. B. also the article G. Gabriel: Art. Linguistic Analysis , in: Historical Dictionary of Philosophy , Vol. 9, 1435–37, which deals with a few selected philosophical-historical contexts and authors .
  4. ^ Words - Die kleine Verräter on zeit.de, accessed on June 22, 2015
  5. ↑ https://www.tagesspiegel.de/wirtschaft/kuenstliche-intellektiven-der-algorithmus-kann-42-dimensions-einer-persoenlichkeit-messen/22756300.html
  6. ↑ https://www.wirtschaftspsychologie-aktuell.de/fachbuch/20180425-klaus-stulle-psychologische-diagnostik-durch-sprachanalyse.html
  7. ↑ https://www.haufe.de/personal/hr-management/eignungsdiagnostik-warnung-von-prof-heinz-schuler_80_457502.html "
  8. ↑ https://www.haufe.de/personal/hr-management/sprachanalyse-eine-neue- Methode-der- personalwahl_80_453994.html
  9. ↑ https://blog.gwup.net/2018/07/12/video-was-verraet-die-sprache-ueber-einen-menschen-mit-prof-uwe-peter-kanning/
  10. ↑ Rena Tangens : Laudation Communication: Precire Technologies GmbH. In: bigbrotherawards.de. June 8, 2019, accessed June 21, 2019 .
  11. ^ On the staff: Top five benefits of speech analytics for the call center (English) . In: TechTarget , February 1, 2009. Retrieved December 28, 2016. 
  12. ^ To the staff: Speech and text analysis (German) . In: TechTarget , January 1, 2016. Retrieved December 28, 2016. 
  13. ^ Larry Skowronek: Do Speech Analytics Tools Change Agent Behavior? (English) . In: ICMI , April 15, 2015. Retrieved December 28, 2016. 
  14. ↑ Jeffrey Fotta: Reverse a Pattern of Poor Sales With Speech Analytics (English) . In: Entrepreneur , January 13, 2015. Retrieved December 28, 2016.