Observing language change and language processing: old manuscripts, new brains
by Professor Aditi Lahiri FBA
5 May 2015
Language is fluid and ever-changing – it varies across time, between speakers and within the same speaker. However, language cannot change in just any way. It is restricted by how language is structured and represented in our brains, specifically in the mental lexicon. An approach which deals with several kinds of variation by combining the study of historical change with the experimental investigation of language processing in today’s speakers may reveal new insights into the nature of language which have previously eluded researchers.
Language change is occurring all the time and while it is often subtle it cannot be halted. There have been letters to the editors of English newspapers about the ‘misuse’ of the English language (especially by the younger generations) for centuries. This preoccupation with language change is not unique to English, nor is it a modern phenomenon: Cicero, writing in 46 BC, lamented the deterioration in the speech of public figures compared with a century earlier.
Lexical change, such as the use of literally to mean ‘figuratively’, is unavoidable and is part and parcel of language as a communicative tool which adapts to changes in societies. Language change is an active and organic process and any language’s speakers are the (unwitting) proponents of this change. Changes first occur in one generation and then some of these changes are taken on by the next new generation of speakers and thus become part of the accepted standard.
Early evidence can sometimes be gleaned from manuscripts. In early scripts, where the orthography was still being established, we can sometimes see alternations of words which correctly reflect the way in which they were pronounced and the general patterns of alternations can be deduced. In the Old High German manuscript written by Notker Labeo (12th century) we see that the word beginnen ‘to begin’ is written with a
or a. Close scrutiny shows that the initial letter depends on the sound of the previous word.
Lexical changes are frequently most noticeable but we also see more subtle changes in the pronunciation and sounds of a language (phonology), that can eventually lead to changes in the way sounds and words are stored in our mental lexicon. An example of this would be the pronunciation of words like car and hair. There are dialects of English where the /r/ sound at the end of a syllable is still pronounced (as is the case in most American and Irish/Scottish dialects) while other English dialects have lost the /r/ in pronunciation. Consider the pronunciation of the words car and hair in Southern British English dialects; while /r/ is retained in the spelling, it is no longer pronounced.
Speech requires a speaker and a listener, and both have their roles to play in language transmission and change. As speakers our job is to try to convey our message to the listener and as listeners we have to make sense of what is being said. However, as listeners we often find ourselves faced with a difficult task since what we hear varies with every speaker and is often distorted by factors such as background noise. There are many different variables, for example, speaker-specific variables like gender, dialect and age of the speaker, the rate of speech, or the noise level of the environment. However, the greatest degree of variability is found in the acoustics of speech itself. No word is ever pronounced in exactly the same way twice – not even by the same speaker, and yet we are able to comprehend speech with remarkable ease.
The answers to questions such as how we deal with the variability of speech during processing and how we access words and their meanings in our mental lexicon ultimately give us insights into how speech is stored in the brain. There are questions concerning whether we store every single token of a word that we have ever heard or whether we take a sparser approach and only store what is not predictable through the application of rules. Looking at the past tense of English verbs, for example, we may store the past tense of all verbs in English as well as the present tense – play, played; wash, washed. Alternatively, we may store only the past tense of irregular verbs, such as go, went and eat, ate, where the past tense is unpredictable, since we understand the rule for forming the past tense of regular verbs.
To gain a comprehensive idea of how the human language processing system works and how language changes, we must draw on separate pieces of evidence from different research areas. It is very much a case of taking individual pieces of evidence from different languages and combining them to arrive at a better understanding of the more general principles that govern speech processing and their relationship with principles of language change.
Aditi Lahiri obtained her doctorate degrees from the University of Calcutta and Brown University. After teaching at UCLA and UC Santa Cruz, she became a research scientist at the Max Planck Institute for Psycholinguistics, in the Netherlands, followed by the Chair of General Linguistics at the University of Konstanz, Germany. Her honours and awards include the Leibniz Prize (from the German Research Foundation, 2000), Fellow of the British Academy (2010), and an honorary life member of the Linguistic Society of America (2013).
© Picture 1: Jessica Emmett (2014)
Pictures 2 & 3: Extracts from De nuptiis Philologiae et Mercurii Codex Sangallensis 872, by Martianus Capella, translated by Notker der Deutsche, 11th century. Pictures taken by Henning Reetz by kind permission of the Stiftsbibliothek, St Gallen (see Lahiri, Aditi & Astrid Kraehenmann (2004)). On maintaining and extending contrasts: Notker’s Anlautgesetz. Transactions of the Philological Society, vol 102(1) 1-55.)