The Evolution Of Grammar, by Joan Bybee, et al | Literature Review

This is definitely a book that can be expected to be assigned in any linguistics course, but it was worth it for the purposes of my conlanging research.


The writers of this book make the case that the analysis of grammar–specifically the grammatical morphemes–through a diachronic perspective. The method of research that they used was to take languages from separate phyla, and compared the grams within them.

Joan Bybee, revere Perkins, and William Pagliuca

Before collaborating as one collective author, they had already published works that they frequently reference throughout this book. They make frequent references to their past research, however they do not use we rather they simply cite their research the same way they do with the other scholars. They had improved on being able to test their hypotheses than when they previously did so in 1991.

Bybee and Pagliuca, specifically, have previously argued that words like will, shall, and be going to retained their original usages, their meanings have changed.


Diachrony is a major theme of the book, since it ultimately showcases the stages of progression from separate phoneme to an affix through phonological reduction. Not only that, but there is also the progression from one specific meaning to a abstract, generalized meaning. As such, in the middle of the book, the authors actually get into detail about the words that were typically used in the ancient forms of the languages until they ended up generalizing. Those words include “to finish” (when using past tense) and “to have” (when referring to a recurring action); as well as “to be” and “to come.” Of course, this opens up a new discussion about how the word “to finish” could refer to a completed action or an action that recurs from the past to the present.

While they argue that unidirectionality is a recurring theme in many if not all languages, they further noted that not all grammaticizations are the same. The usage of affixes in other languages are different from others. In Romance languages, the words “to have” and “to be” with the past participle construction developed into the perfect case in Spanish; while the spoken perfective in French developed from an older perfective inflection. Sometimes, instead of the affixations of words, the opposite might occur if the grammar becomes so complicated that periphrasis might be used.

There are also differences in the the resultative and the completive stages. While completives result from dynamic verbs, since they result in change and movement in a situation; resultatives spring from static verbs, because resultatives usually apply to intransitive verbs.

There is a lot of importance of cognitive and communicative contexts behind grams, since they ultimately shape their new formations. They do not provide an entirely new meaning, rather ones that they inherit through the social context. As such, this would involve the use of temporality when dealing with that transitioning stage. The use of space and time becomes distinguished when there is a predicate involved, specifically detailing the situation that is happening. If it happened in the past but is continuing, then an imperfective would be used, just like in the English language; but if the action was completed, then the use of simple past would be used; on and on.

There is also the issue of the perspective of the speaker, specifically when taking into account the length of the remoteness of time. There are some grams that focus on the predicate a long time ago, a few days ago, yesterday, or any point in time.

Historical Context

The only mentions of history occur with the mentions of languages as examples of grammaticizations. In the case of Modern Irish, it uses a separate pronoun in VSO order, whereas in previous stages of Irish Gaelic, the pronouns were affixed to end of the verb in the same sentence order. The authors also took the time to tie together the affixes of the Bantu languages with a single reconstructed Proto-Bantu word.

Creole and pidgin languages are given a special focus, because it is where the grammaticizations take place in real time. Although, creoles and pidgins are considered their own category since they follow the same rules that can be expected in any of them.

As for the authors themselves, they mainly wrote this book in order to add to the critical conversation concerning the meaning labels behind verbal grams, which had been a fairly recent issue at the time this book was written. They also sought to argue against positions taken on grams, such as arguing that the Time-Is-Space is not valid, since the use of space in grams do not pertain to any specific location. Another issue that they had was when the previously held belief about the realis/irrealis modalities was that they were strict distinctions, whereas the authors simply indicate that these modalities can convey a multitude of points of time.


When the writers starting discussing the use of space in language, it made me think about Guy Deutscher, who made the case in his book that language originally focused on space while it gradually expanded into more abstract meanings. This is especially important since it is evident in languages today, such as Russian which employed their locative affix to their verbs in order to create a perfect tense. As such, the generalization of this new gram is to indicate that the verb is pointing to a limit.

And of course, humans are unique in that they can develop abstract imaginations, as opposed to other primate species. This would be further pinpointed by Yuval Noah Harari when examining the whole of human evolutionary history. The authors would definitely proceed where Harari did not in his book.

When it came to the nuances in human interaction in the Modality chapter, it definitely made me think about Mark Rosenfelder’s book about conlanging in which he gets into extraordinary detail about those very nuances.

As well as those nuances, it also made me understand that every language can always make do when they do not have a particular word or affix to describe something. Biblaridion demonstrated this in his video on the worst conlang he ever did. He then talked about how Latin is able to use the word “and” in such a way as not to involve a completely different affix. This enables a conlang would appear natural, not unnecessarily complicated.

Writing Style

As expected of any academic work, there is a lot of dense language with large terminologies. The authors would assume that the reader knows what they are writing about and are willing to make their case to them about how grammaticization is diachronic.

  • Gram: Grammatical Morpheme
  • Diachrony: the chronology of any language development
  • Generalization: the transition from specified usage to generalized usage.
  • Grammaticization: the process of transitioning from one stage of phonology and meaning to another.
  • Unidirectionality: the process of grammaticization through a single sequence. Since it can literally mean “a state of one direction,” then it could help create a mnemonic that would be easy to remember.

It also involves looking at phonemes both by their meanings and their contextual usages. In the case of will, shall, and be going to, they appear the same but are used differently. Will involves the desire to act; while shall involves both the speaker’s and recipient’s participation; and be going to is an odd combination of different generalizations, since going to involves the grammaticization of to go which instead of referring to a specific location, it refers to the completion of an act. That phrase would go on to proceed through phonological reduction by contracting into the informal phoneme gonna.

The authors also make it clear that the grams’ meanings do not devolve or become destroyed, rather they erode. They use this type of word choice throughout the book, so as to clarify that the meanings of independent words in proto-languages shift from physical to abstract, either in the case of aspectual or temporal meanings. Simply put, they do not disappear, rather they take new form.

Real-World Application

This book is not to be read only once, rather it is basically a reference guide for any linguistic research. The important point that can be made is that certain words/affixes do not emerge ex nihilo, rather they evolve from previous words within the span of many millenia.

If anyone is intimidated by the dry, dense language throughout this book, the only assurance that I can offer is that the authors provide a lot of graphs and they consistently make the case that almost all languages follow the same pattern. So, it is important to keep unidirectionality in mind when looking not just at this book, but at all of the languages that are spoken in this world. Once you start learning new languages, it becomes easier since you can notice the patterns–the protyposymbiosis if you will.

Since the authors of this book made the argument that the grammaticization from separate phonemes to grams is universal in all languages, I can see why Biblaridion would suggest it. This provides an overview of the functions of any language. It would primarily be helpful to those who are establishing language families for their conlangs.

Recommend This To…

  • Anyone who has a basic level of understanding about linguistics. For anyone else, it would appear daunting, especially with the heavy usage of linguistic terminology. All I can really say is that it is easier to know your basic Greco-Latin affixes.
  • Anyone who wants to study another language in an advanced level. It can be paradoxical to imply that reading this dense book will make any language-learning experience easier, though that is the case with this book.


Image Attribution: Micky Milkyway

Support me on Patreon.

Leave a Reply

Your email address will not be published. Required fields are marked *