At the end of 2023, several high-profile dictionaries published their ‘Word of the Year’. They are what the Oxford English Dictionary (OED) describes as “a word or expression that has attracted a great deal of interest over the last 12 months” and which may have “lasting potential as a word of cultural significance” [1]. Predictably, there was fallout.

On the matter of ‘rizz’ (an abbreviation of ‘charisma’, popularised by Gen Z) winning the OED’s 2023 crown, Angus Colwell wrote in the Spectator that “it’s dispiriting the way the internet is hollowing out language” and—proving that a lack of understanding about cognition and language is no barrier to opinions on either—”If we speak worse, we think worse” [2].

Amongst the proletariat, comments on BBC News’ Instagram post about the new entry into the language included such objections as [3]:

• “William Shakespeare is truly shaking rn [right now]”;
• “Seriously I’m too old for this”; and
• “We should’ve stopped breeding in 2004”.
  
  

What Colwell’s article and each of those Instagram comments object to, with varying levels of indignation, is the idea that language evolves organically rather than systematically. Moreover, dictionaries, being permissive of such an idea, are complicit in this aberration. There are real words and feeble imitations, and dictionaries ought to know the difference.

This reflects a common misconception—undoubtedly informed by heated Scrabble disagreements—that dictionaries are the final word on correct usage: the prescriptivist’s holy text. Words in the dictionary ought to be inherently credible (especially when a triple word score is concerned) and uncredible ones, like ‘rizz’, have no business being admitted to them. Indeed, this is the likely function of dictionaries for many of us: rarely does pettiness feel quite so vindicated as discovering your preferred pronunciation or spelling is also the dictionary’s.

However, this is not a dictionary’s intended purpose. Language is ever-changing: a metamorphosis driven by cultural shifts, technological advancements, and social trends. It’s why we no longer use Old English ‘cyningdōm’ and instead use ‘kingdom’, and say ‘the’ instead of ‘se’. Language is similar to a living organism: adapting, absorbing, and shedding words. It is the function of dictionaries not to judge these changes but to observe and document them.

This blog explores that responsibility by examining the intersection between language change and dictionaries: the necessary trade-offs the latter must accept to record contemporary usage, and the methods employed to do so. What we will find is that dictionaries, while imperfect, remain a unique reservoir for our language.

Language Change and the Limits of Dictionaries

The English language is vast, rich, and labyrinthine. Centuries of constant change—from invasions which brought multitudes of new lexis and plagues that ruptured linguistic power bases—have rendered the task of immortalising the sheer variety of English in a single text near impossible.

Thanks to events like the ‘Great Vowel Shift’—a period of sound change over several hundred years, culminating at the beginning of the eighteenth century—the pronunciation of certain vowels transformed, intensifying divergence between southern and northern varieties of English [4]. ‘Sheep’, for example, was pronounced like ‘shape’ in Middle English, and ‘doubt’ like ‘doot’ (similar to the long ‘oo’ in ‘soup’). Many varieties of English have retained approximations of these older expressions. But in dictionaries, such as the OED, only the shifted pronunciations, common to the south-east of England, are recorded. ‘Doubt’, for example, has its listed pronunciation in British English solely as /daʊt/ (pronounced as ‘dowt’) [5].

This is not because there are distinctly correct and incorrect pronunciations but because of historical interplays between power, aspiration, and geography.

Standardisation: A Brief History of Power in the English Language

Standardisation—a set of rules governing language use in particular contexts—is not an overnight process, and is a relatively recent phenomenon for British English as a whole. For much of the language’s history, people have spoken a variety of dialects that were more likely to reflect their regional roots than social class or level of education.

Before standardisation, regional accents rarely impeded ascension to power: in the seventeenth century, Francis Drake and Walter Raleigh are likely to have had Devonshire accents, whilst James VI of Scotland and I of England would have spoken English with sounds and lexis derived from Scots (a distinct Germanic language). Moreover, it was common to spell words as one pronounced them, leading to incredible levels of spelling variation in historical texts: ‘through’ had at least 50 [6].

London, however, and more generally the south-east of England, had been amassing influence for centuries. Preferred to the first Roman capital, Colchester, and the Saxon seat of Winchester, its centralised role in England—and then Britain—was cemented when William the Conqueror made it his capital following victory at the Battle of Hastings. This geographic concentration of power continued as developments in language technology and documentation evolved. Caxton’s introduction of the printing press 1476 and the 1755 publication of Samuel Johnson’s seminal Dictionary of the English Language allowed for more influential variants in English to be elevated above others.

Moreover, events like the Bubonic Plague had the inadvertent consequence of decimating the Norman ruling nobility, which allowed an English middle class to flourish. This meant that by the seventeenth century, social mobility became more viable. However, achieving an upward social trajectory, required one to imitate the linguistic varieties of society’s most influential: London’s upper class. Northern pronunciation of ‘Bath’ (with a short ‘a’) became ‘bahth’ with a long one. The Cockney dropping of ‘h’ in words like ‘house’ was eschewed in favour of full enunciation. In short, the plot of My Fair Lady (1964).

These intersecting events would result in the English variety we now call Received Pronunciation (RP): a sociolect that reflected social class, aspiration, and level of education, rather than regional origin. In modern times, we often refer to it as BBC English or the Queen’s English.

The OED’s Model

The OED roots its listed pronunciations in RP. Crucially, however, this decision is described as pragmatic rather than righteous. They state that “RP for OED is not concerned with prescribing ‘correctness’” but rather its broad intelligibility to a spectrum of native English speakers [7]. When describing how to use the dictionary, the OED makes three points clear:

• RP is only a single variety of English, and the OED cannot hope to record and present all feasible varieties of pronunciation (or, indeed, spellings);
• While more prevalent in south-east England, RP is not regionally-bound; and
• There remains a variety within RP that can be informed by regionality.
  
  

All of this is to say that the OED is doing the best it can with finite resources and following the path of least resistance. It attempts to capture large volumes of language data in a way that does not alienate certain groups of speakers or invalidate their usage. Inclusion in the dictionary—as is the case for Cambridge, Merriam-Webster, and Collins—is empirically based on the volume of recorded usage.

So, What is a Dictionary?

Dictionaries such as the OED, Cambridge, Merriam-Webster, and Collins are maps, not manuals: they record rather than instruct. They are cameras repeatedly snapping the English language as it moves and shifts and blurs, is made and remade by invention, change, and obsolescence.

Language is a social phenomenon, and so change—powered by generational, geographic, and technological upheaval—is inevitable. One need only look at the etymological history of ‘nice’ in the OED to chart the evolution of its meaning—from ‘foolish’ and ‘lasciviousness’ to ‘extravagant’ and finally ‘pleasant’—and realise language has never been fixed [8]. Word of the Year contenders represent only a minority of new additions to the language recorded in dictionaries: the OED added 71 new terms in September 2023 alone [9].

Dictionaries are valuable but that value needs to be contextualised to be understood. They can tell us the source of new words—increasingly online spaces like TikTok or Reddit—and whether words are slipping towards archaicism and antiquity, or whether we might now only find them in literature, rather than spoken discourse. For learners of the English language, dictionaries can be an important repository of formal lexis e.g. used in academic or official documents. But dictionaries only ever describe the English language as it is and not how it ought to be. As a result, they are a critical historical record of language: its innovation, adaptation, and even death.

Perhaps at the end of 2024, when a host of new contenders for Word of the Year are announced to outrage and vexation, we might be mindful of what Samual Johnson wrote in the preface to his 1755 dictionary: “sounds are too volatile and subtle for legal restraints; to enchain syllables, and to lash the wind, are equally the undertakings of pride, unwilling to measure its desires by its strengths.”