‘Slaves’ and ‘Slave Owners’ or ‘Enslaved People’ and ‘Enslavers’?

Abstract

Studies of slavery increasingly refer to ‘enslaved people’ rather than ‘slaves’, and, to a lesser extent, to ‘enslavers’ rather than ‘slave owners’. This trend began with scholarship in the United States on plantation slavery but has spread to other academic publications. Yet ‘slave’ continues to be widely used, indicating not everyone is aware of the change or agrees with it. Despite this, few historians have justified their terminology. After surveying the extent of the preference for ‘enslaved person’, I discuss arguments for and against it. Supporters of using ‘enslaved person’ argue that this term emphasises that a person was forced into slavery – but this emphasis means it is less able to accommodate early medieval cases where people sold themselves into slavery. The accompanying preference for ‘enslaver’ over ‘master’ obscures dynamics of ownership and manumission. In addition, ‘enslaved people’ and ‘enslaver’ do not necessarily bring us away from the perspective of slaveholders to the perspective of slaves. Nor are they essential for readers to appreciate the humanity of slaves. Overall, historians should use this issue as an opportunity to reflect on the extent to which scholarship of transatlantic slavery should set the terms of debate for slavery studies in general.

In the 2022 volume of this journal, of the four articles that mentioned slavery, two referred to ‘slaves’, while two referred overwhelmingly to variants of ‘enslaved people’.¹ The authors of one of the latter articles also used ‘enslavers’ as its main term for slave owners.² These discrepancies highlight how ‘enslaved people’ and ‘enslavers’ have grown in popularity as replacement terms for ‘slaves’ and ‘slave owners’ respectively, but have not yet become standard worldwide. This is unsurprising, given that the debate over what terms to use when writing about slavery has taken place rather discreetly, in online forums, with very little published discussion of the theoretical issues at stake. In 2018, Laura Rosanne Adderley, a historian of the nineteenth century, tweeted that, ‘“Enslaved” solves some problems, but may create others.’³ Adderley herself still prefers to use ‘enslaved people’, but her tweet was an acknowledgement from proponents that this term may not always be appropriate. However, several years later, problems with using the term ‘enslaved people’ remain under-discussed. This is not just a matter of linguistic niceties. At their best, changes in terminology can encourage us to rethink our assumptions and stimulate better historical analysis, as happened with the shift away from using ‘feudal’ to describe the power structures of medieval Europe. But, unlike ‘feudal’, where the problems with the term have long been established, an academic debate over the relative merits of ‘slave’ and ‘enslaved person’ has barely started. This article is intended, therefore, to develop the discussion, rather than end it.

Historians need to think carefully about what terms we use – including those of us working on societies before AD 1500. There have been forms of slavery across many different periods and places, from medieval Korea to pre-Columbian America.⁴ Approaching the debate from my own research interests in early medieval Europe, I find ‘enslaved person’ and ‘enslaver’ to be inadequate substitutes for ‘slave’ and ‘slave owner’. I appreciate that other historians, especially those working on different periods to me, may come to different conclusions – but that is precisely why more academic discussion is needed. I hope, indeed, that this article will inspire interesting responses from historians who disagree with me. Once this debate gets going, it should clarify the strengths and weaknesses of using the term ‘enslaved person’ when writing about slavery in different historical contexts.

Let us start by noting the origins and extent of the preference for ‘enslaved people’. Back in 1999, Deborah Gray White stated that, were she to rewrite her book on female slaves on plantations in the American south, she would refer to them as ‘enslaved’ rather than ‘slaves’, because, in her view, ‘“enslaved” says more about black people without unwittingly describing the sum total of who they were’.⁵ But it is only more recently that ‘enslaved people’ has gained widespread momentum. Two important works laid some of the theoretical groundwork favourable to this trend: first, a 2003 article by Walter Johnson, which called for historians to speak of ‘enslaved humanity’ rather than ‘slave agency’, in order to ‘imagine a history of slavery which sees the lives of enslaved people as powerfully conditioned by, though not reducible to, their slavery’;⁶ second, Joseph Miller's 2012 book, The Problem of Slavery as History, which argued that a focus on slavery as an abstract institution obscures the importance of slaving as a dynamic and ongoing process.⁷ However, both Johnson and Miller still referred to ‘slaves’, which suggests that they thought that their respective goals – affirming the personhood of slaves and emphasising that slavery was not a static institution – were achievable without wholesale adoption of the term ‘enslaved people’.⁸ Indeed, advocates of the term ‘enslaved people’ more often cite a rubric which Daina Ramey Berry developed for a 2012 encyclopedia on enslaved women in America, alongside P. Gabrielle Foreman's community-sourced style guide on writing about slavery, both of which also recommend referring to ‘enslavers’ rather than ‘masters’.⁹ The decision to eschew ‘slaves’ by the editors of the New York Times's high-profile ‘1619 Project’, on the legacy of slavery in America, has likewise been influential in encouraging academics, heritage institutions and media outlets to follow suit.¹⁰ Works by Thomas A. Foster and Daive A. Dunkley are among academic publications on American plantation slavery which use the term ‘enslaved person’ except when paraphrasing or quoting from primary sources.¹¹

Foreman explicitly states that her style guide is intended as a series of suggestions rather than an orthodoxy to be enforced, given the ‘particularities of institutions of slavery in various parts of the Americas, Europe, Africa, and Asia, and also considering how slavery changed over time’.¹² Yet, noticeably since 2019, the preference for ‘enslaved person’ has grown beyond American scholarship of American slavery. In the United Kingdom, educational resources for the worlds of both the Roman and British Empire (notably BBC Bitesize and the Cambridge Latin Course) have started using ‘enslaved person’ instead of ‘slave’.¹³ Nonetheless, ‘enslaved person’ has not percolated everywhere. Only two of the eight research articles in the first 2023 issue of the Slavery & Abolition journal primarily referred to ‘enslaved people’ rather than ‘slaves’.¹⁴ Recent books on aspects of the British Empire by Linda Colley and P. J. Marshall have used the term ‘slaves’ outside of direct quotation of primary sources.¹⁵ Contributions to a 2023 edited book on the Dutch and Portuguese empires in South America also interchangeably refer to ‘slaves’ and ‘enslaved Africans’, though some of the authors appear to prefer the latter term.¹⁶

‘Slave’ remains common in both US and non-US academic publications dealing with the ancient world – for example, in books by Franco Luciani, Peter Thonemann and Michael Flexsenhar III.¹⁷ However, other historians, like Roberta Stewart, have started referring to ‘enslaved people’ instead.¹⁸ A recent book on slavery and sexuality in classical antiquity is illustrative: the editors, Deborah Kamen and C. W. Marshall, avoided using ‘slave’ in their introduction, as do around half of their fellow chapter authors – but the other half still used it.¹⁹ Perhaps this divide reflects concerns among ancient historians about letting scholarship of early modern slavery (literally) set the terms of debate around slavery in their period, which alerts us to the possibility already that linguistic models inspired by one historical context may not be appropriate for another. Interestingly situated on the spectrum of this debate among classical historians is Kostas Vlassopoulos, whose 2021 book aligns with proponents of ‘enslaved persons’ in arguing that the term ‘slaves’ can hide the multiple identities available to the unfree, whereas it nonetheless refers to ‘slaves’ throughout.²⁰ Vlassopoulos emphasises that he is not calling for a change in labels, but for historians to avoid thinking about people as merely and only as slaves. I will return to his work later. For now, I shall proceed with my survey.

Given the ongoing ubiquity of ‘slave’ in scholarship of other periods, it is unsurprising that prominent recent works in the field of early medieval slavery continue to use ‘slave’. Alice Rio, in her 2017 monograph, though she primarily spoke of ‘unfree people’ and ‘unfree status’ as a way to acknowledge the overlap between forms of bondage in this period, still used ‘slave’ to refer to those on the ‘most heavily subjected end of the spectrum of unfreedom’.²¹ As with the previous volumes in the series, the chapters of the Cambridge World History of Slavery: Volume 2, AD 500–AD 1420 referred to ‘slaves’, though some of the contributors more typically – but not exclusively – referred to ‘enslaved people’.²² Chris L. de Wet and all but one of his fellow authors routinely used the term ‘slaves’ in Slavery in the Late Antique World (2022).²³ As do Mary E. Sommar, Roy Flechner and Janel Fontaine, despite all publishing after 2019.²⁴ There are exceptions. For example, Ben Raffield's work on the Viking age has eschewed references to ‘slaves’, though this partly reflects his focus on the process of enslavement via raiding, as well as the difficulty in establishing whether captives were enslaved permanently, ransomed, killed or met another fate.²⁵ But, overall, the difficulties of translating servus and the issue of slavery versus serfdom continue to be the greater source of controversy when it comes to writing of ‘slaves’ in early medieval Europe.²⁶

Constrained by lacking an alternative label, medieval historians have also taken care that their use of the word ‘slave’ does not cause their readers to equate the conditions of unfreedom in the Middle Ages with the racialised chattel and plantation slavery of later periods. It would almost be fortuitous, then, if it were not a partial and probably temporary development, that historians working on American slavery have, by being quicker to change their terms than those of earlier periods, incidentally created linguistic distance between ancient and medieval ‘slaves’ and early modern ‘enslaved Africans’. Indeed, the divergence between forms of slavery may well merit a divergence in terms. Below, I will set out objections to using the term ‘enslaved people’ that historians of plantation and transatlantic slavery could disregard as irrelevant to their periods. Undoubtedly, these scholars know a lot more about slavery in America than I do. It is not my intention to disregard their expertise, nor the particular significance this issue might have for those whose recent ancestors were enslaved, who may understandably see an emphasis on the personhood of enslaved Africans as important in the context of anti-Black racism. But if historians studying different forms of slavery can come to different conclusions about which terms to use, then that gives me even more reason to make my case. By thinking about terms, we can clarify quite how complex and varied slaveholding practices have been.

Despite the limited acceptance of the term ‘enslaved people’ indicating that not everyone agrees it is a necessary or appropriate substitute for ‘slaves’, there is relatively little academic literature devoted to either arguing in favour of it or against it. Back in 2006, David Brion Davis criticised the term ‘enslaved people’ as inelegant and euphemistic, but only in a footnote.²⁷ Alluding to instances such as this, Berry regretted in 2012 that debates on the terminology of slavery were ‘hardly discussed, or equally unfortunate, were left buried and hidden in the footnotes’.²⁸ Outside of a few honourable exceptions, this continues to be the case.²⁹ While online forums and conferences have given historians an opportunity to discuss terms with each other, this has mostly not translated into academic publications. Considering that the movement towards ‘enslaved people’ largely appears to have been driven by individual scholars choosing to adopt the term, rather than journals or publishers, it is surprising that historians have been so unforthcoming about their motives – though perhaps if editors did have set policies, that would spur some into action. The lack of transparent reasoning by historians working on societies beyond the early modern North Atlantic world is especially disappointing, given Foreman's invitation to engage with the historical particularities of slavery. However, some historians – by no means all – have referred to Foreman's style guide, acknowledging their debt to her guidance.³⁰ That even ancient historians have cited this style guide, with its recommendations inspired by scholarship on transatlantic slavery, is testament to the clarity, cogency and accessibility of Foreman's arguments. But it also highlights the poverty of extended academic analysis of this issue.

Readers of historians who have not elaborated on their preference for ‘enslaved people’ are forced to infer their motives from what other advocates of the term have stated online, often outside of formal academic forums. Even approached charitably, this process of inference is necessarily fraught. It is very likely that there are arguments for using ‘enslaved people’ which are not found in the standard reference pieces that largely form the basis of my citations.³¹ Indeed, some historians who use ‘enslaved people’ have not followed Foreman's style guide in other respects, indicating that these historians have a slightly different approach to the terminology of slavery. Berry and Foreman's advice that slave owners be referred to as ‘enslavers’ is the most notable and popular of their secondary recommendations – and so will be the subject of much of the ensuing discussion – but even this change has still only partially been adopted by adherents of ‘enslaved people’.³² I therefore encourage academics who believe that I have not recognised all the advantages of using ‘enslaved people’, or that the proponents I cite do not represent their views, to set out their arguments in response to this article.

I shall nonetheless address a range of reasons for using the terms ‘enslaved person’ and ‘enslaver’ by discussing, in turn, the historical, ethical and political issues which the shift towards them raises: how far are these terms historically accurate; how far do they help us to appreciate the humanity and perspective of the unfree, and thereby avoid the harmful implications which advocates argue that ‘slave’ and ‘master’ possess; and should historians see any professional advantages in using these upcoming terms in the context of faculty politics? I find that significant problems, arising especially out of the complicated circumstances of early medieval slavery, prevent ‘enslaved people’ from fulfilling its potential to bring historical and moral clarity to slavery studies. Nonetheless, ‘enslaved people’ lacks the fundamental flaws of the accompanying preference for ‘enslaver’, which obscures more than it illuminates. I finish by recommending that historians of all periods engage with this debate more than they have hitherto done.

Historical considerations

The most straightforward argument for referring to ‘enslaved people’ rather than ‘slaves’ is that it brings historical clarity. Berry argued that, when discussing ‘people of African descent held against their will’, ‘“enslaved” emphasizes the reality that enslavement was an action – a verb enacted on individual(s) rather than a noun, “slave,” that describes a social position these individuals presumably accepted’.³³ Foreman's style guide puts this argument succinctly: ‘People weren't slaves; they were enslaved.’³⁴ But how far is this applicable to all those who were ever in slavery? The fourth- to fifth-century bishop of Helenopolis in Bithynia, Palladius, wrote about an Egyptian slave who deliberately sold himself into slavery – twice – because he had vowed a life of asceticism and wanted to convert his prospective owners to Christianity.³⁵ Even if we suspect that most people did not see slavery as an opportunity for piety or that Palladius fabricated these events, there is good evidence that cases of people voluntarily becoming slaves were unexceptional in the first millennium. Despite the laws against selling yourself as a slave in the Abbasid caliphate, a newly discovered ninth- or tenth-century papyrus shows the author, a prisoner, contemplating doing just that, partly so he could escape his living conditions, but also perhaps as a way to dissolve a relationship with an unhelpful patron.³⁶ Meanwhile, Alice Rio has drawn attention to the number of early medieval European cases in which people sold themselves into servitude.³⁷ Moreover, she has shown that self-sales were not necessarily the result of direct coercion or even poverty: the variety of terms and conditions found in legal formulae indicate people of moderate wealth choosing to enter unfree service in exchange for protection or another perceived benefit from their master.³⁸ In her words, ‘People who sold themselves were not always passive victims and could be quite shrewd in bargaining over their freedom.’³⁹

Those who want to explicitly mention the personhood of the unfree could practically accommodate these early medieval examples using the term ‘self-enslaved people’. Some may see this phrase as a linguistic contortion, while others may believe that it possesses a certain stark elegance. Either way, we are still left with something of a paradox. As noted above, Berry and Foreman suggested that ‘enslaved people’ should be preferred because it intuitively implies that people did not accept their slave status. Therefore, to apply the label ‘enslaved people’ to those who sold themselves requires us to negate an implication which is meant to be a key source of that label's strength. Of course, ‘enslaved people’ is not misleading when people became slaves as a result of conquest or raiding. The particular difficulty for the early medieval historian is that it is rarely obvious from our extant sources where a slave came from. It can therefore be advantageous to use the term ‘slave’ precisely because it does not specify exactly how a person lost their free status.

If someone is an ‘enslaved person’, it follows that they had an enslaver. This enslaver could have been a human trafficker, a raider in a war band, a judge, prosecutor or lawmaker who prescribed penal enslavement as punishment for a crime, and/or, as we have seen, the slave him- or herself. But it is actual slave owners that Foreman and Berry have emphasised should be called ‘enslavers’.⁴⁰ Like ‘enslaved person’, the label of ‘enslaver’ reflects a growing perspective that slavery was a dynamic process in which masters continually and violently reasserted the slave status of those they claimed as property – and, indeed, created that status for those born to slaves already under their ownership. Certainly, the response of slaveholders to fugitive slaves in late antiquity brings into focus the need of owners to constantly re-enslave people to maintain their control. The sixth-century bishop Gregory of Tours wrote about how a man named Leudast, whose father Leucadius had also been a slave, fled his owners repeatedly, until his owners mutilated his ear, implicitly so Leudast would understand that others would recognise him as a slave whenever he tried to escape.⁴¹ We can interpret the mutilation and recaptures of Leudast as representing only the most extreme and visible end of the recreations of slave status which slaveholders practised on their human property. Gregory of Tours reviled Leudast, and some historians are sceptical of his version of events.⁴² Still, it highlights how contemporary slave owners might have acted as enslavers, and so, by extension, how even those born into slavery can be considered ‘enslaved people’.

Nevertheless, several important objections remain to making ‘enslavers’ the standard term for slaveholders. Although slave owners used strategies of enslavement, enslavement is not the only dynamic relevant to slaveholding. ‘Enslavers’ ignores the opposite direction to enslavement in which slaveholders sometimes acted in relation to their slaves: manumission. It is easy to be cynical about instances of owners freeing slaves, but, as Ilaria L. E. Ramelli and others have shown for parts of the late antique and early medieval world, manumission could be understood as an act of Christian charity pertinent to the salvation of both master and slave; writings about slavery should be able to convey this dynamic.⁴³ Furthermore, referring to slaveholders as ‘enslavers’ elides the difference between those who first forced a person into slavery, and the people who subsequently owned slaves as a result of a sale, gift or inheritance. There was plenty of overlap between these groups in both theory and practice, but the distinction between slave owners and enslavers mattered in late antique society. It meant Augustine could justify slavery theologically while condemning slave merchants who kidnapped people to sell them.⁴⁴ It meant a bishop like Caesarius of Arles could believe in the virtue of ransoming captives who faced enslavement, while owning slaves himself.⁴⁵

This distinction between slavery and enslavement might have helped early Christian slaveholders assuage their consciences, but it could also have been important to certain slaves themselves. Many enslaved people likely accorded a special violence and significance to the experience of being forcibly separated from their home communities, trafficked away and sold as property. The Confessio of St Patrick, ‘the closest thing we have to a slave narrative from antiquity’, emphasised how, while born to a free father, he was enslaved as a youth: ‘The Lord brought the wrath of his mind upon us and scattered us among many peoples even to the end of the earth.’⁴⁶ It is St Patrick's own self-awareness as an enslaved person who had a life before slavery that directs us to be sensitive to the particular role of enslavement as the initial act by which he was severed from his freedom and family. Though slavery encouraged and perpetuated such acts of enslavement institutionally, it should not be conflated with them. Indeed, that there is an ongoing need to distinguish the process of enslaving from the condition of servitude probably explains why some of those historians who have substituted ‘the enslaved’ for ‘slaves’ have not substituted ‘enslavement’ or Miller's preferred term of ‘slaving’ for ‘slavery’ – even though ‘enslavement’ achieves the same purpose as ‘the enslaved’ in emphasising that people were forced into servitude.⁴⁷ So while the term ‘enslavement’ can be extended to ordinary slaveholding practices, there is value in maintaining its conceptual distance from slavery. This distance cannot be maintained while referring to owners only as ‘enslavers’, but at least ‘enslaved people’ can theoretically operate in relation to a range of possible enslavers.

Perhaps recognising the problems which ‘enslavers’ creates, some adherents of using the term ‘enslaved people’ have used the terms ‘slaveholder’ and ‘slave owner’ while rejecting the term ‘master’ – even though Foreman's style guide advises against all of these terms.⁴⁸ The issues with ‘master’ as a word which implies domination and superiority are easy to grasp, but are surmountable so long as the reader understands that ‘master’ equates to contemporary concepts and legal categories, and should not be taken to mean that someone was naturally superior to those that they owned. Still, there are fewer reasons to defend and retain the term ‘master’ – so long as ‘slaveholder’ and ‘slave owner’ are available as acceptable substitutes.

‘Enslaver’ and ‘enslaved person’ can be used correctly and valuably to emphasise processes of enslavement, when enacted on free people, captives, fugitives and those born to slave families. But neither can work as historically accurate replacements for ‘slaves’ and ‘slave owners’ wholesale. ‘Enslaver’ is too uncompromising in making enslavement the defining dynamic of slaveholding. While the term ‘enslaved people’ has more flexibility, early medieval cases of self-sale strongly bring into question one of the supposed key advantages of using it: that is, to emphasise that slaveholders forcibly imposed unfree status on other people. By contrast, that ‘slaves’ and ‘slaveholders’ do not go into specifics about the origins of a person's unfree status makes it easier for historians to use them when discussing a range of societies.

Ethical considerations

There is more than historical accuracy at stake in this debate. For some historians, using ‘enslaved person’ and ‘enslaver’ appears to be an act of justice which better conveys the humanity and perspective of the unfree than ‘slave’ and ‘master’. Kamen and Marshall have argued that ‘using terms like “slave” and “master” reinforces the enslaver's viewpoint’.⁴⁹ However, I would contend that the term ‘enslaver’ can do that as well. For certain Roman and early medieval warlords, being thought of as an ‘enslaver’, and not just a mere ‘master’, would probably have been a source of pride. In Roman triumphs, victorious generals showed the citizens of Rome that they had defeated and enslaved formerly free barbarian enemies by parading captives in chains before their chariots.⁵⁰ The Emperor Augustus boasted in an inscription recording his accomplishments, the Res Gestae, that during the war in Sicily he returned 30,000 fugitive slaves to their owners for punishment.⁵¹ Gregory of Tours wrote that a king promised his men that they would get many cattle and slaves if they followed him into battle.⁵² Meanwhile, David Wyatt has concluded that, in medieval Scandinavia and the British Isles, ‘abduction and enslavement, particularly of women and the young, were … methods by which rising leaders and their war bands established their status’.⁵³ Therefore, ‘enslaver’ can also reflect the attitudes and aspirations of certain slave owners.

It is also far from clear that ‘enslaved’ and ‘enslaver’ bring us any closer to the viewpoints of slaves. While historians should heed Vlassopoulos's forceful argument that slaves likely did not understand themselves only as being slaves, Vlassopoulos himself acknowledges that ‘most slaves experienced slavery as a direct relationship with their masters’.⁵⁴ Indeed, many slaves may well have considered their owners to be their masters, given the control they exercised over their lives. Crucially, this control often went past the point of manumission, even after an owner's death: the seventh-century will of bishop Bertram of Le Mans obliged his freed slaves to annually gather at his tomb and perform tasks for an abbot.⁵⁵ Furthermore, relationships which went beyond an owner could actually serve to reinforce a person's self-perception as a slave. Ecclesiastical slaves in late antiquity may have conceived of one such relationship as being with God. Lisa Kaaren Bailey has suggested that the contemporary belief that pious devotion involved ‘slavery to God’ could have influenced the self-understanding of slaves at churches and monasteries, to the extent that they accepted their slave status as what God intended for them.⁵⁶ At least that brought the consolation that their service would be rewarded in the hereafter. But as Bailey has acknowledged, our reliance on early medieval texts not written by enslaved people limits our ability to know with any certainty what their views were. The only thing that can be safely assumed is that neither all slaves nor all slave owners would have shared a single, straightforward opinion on the nature of slavery. This observation may seem self-evident, but it needs to be stressed. Someone who was enslaved after a lifetime of holding free status may have had a very different understanding of their situation and own identity than someone who, born into bondage, saw freedom not as a lost status, but a goal to be achieved through manumission. Moreover, in certain contexts, it was not just the attitudes of slaves and slaveholders that probably converged, but their practices. Slaves owned other slaves across the medieval period, creating problems for any attempt to use an ‘enslaver’ and ‘enslaved’ dichotomy to achieve moral clarity.⁵⁷ Ultimately, what is at stake is not the viewpoint of historical enslavers and enslaved people – both are, after all, deceased – but really our own. The rhetorical effect of the labels ‘enslaver’ and ‘enslaved people’, which is to always keep the atrocity that slave owners were committing against other human beings in the forefront of the mind of the reader, draws its power from the abhorrence most people have towards slavery today.

Proponents of referring to ‘enslaved people’ argue that the term ‘slave’ prevents the reader from achieving this same recognition of the humanity and experience of the unfree. Nikole Hannah-Jones of the ‘1619 Project’ claims that ‘The alternative term “enslaved person” accurately conveys the condition without stripping the individual of his or her humanity,’ thereby implying that the label ‘slave’ is dehumanising.⁵⁸ Historian Eric Foner disagrees with this assumption. He has stated, ‘I do not think that “slave” suggests that this is the essence of a person's being.’⁵⁹ But scholarship on ancient and medieval societies supports an awareness that ‘slave’ was not a neutral descriptor but was part of the construction and legitimisation of the system of slavery. For the classical world, Vlassopoulos argues that when certain authorities labelled people as ‘slaves’, they were affirming that that categorisation was ‘all that mattered’ about the lives of those enslaved.⁶⁰ For early medieval Europe, Rio has emphasised that unfree status was ‘the result of an act of labelling and not … a static object’.⁶¹ We may ask, though, if the unconsensual imposition of the label ‘slave’ amid unequal power relations makes it so different from (for example) ‘conscript’ or ‘convict’. Advocates of ‘people-first’ language in general could fairly respond here that we should refer to ‘convicted persons’ or ‘conscripted persons’ as well. Yet when we speak of ‘slave owners’ or ‘enslavers’, or ‘bishops’ or ‘kings’, we take for granted that they were all human beings, whose lives and identities went beyond these labels, without needing to address them as ‘slave-owning people’ or ‘people who were enslavers’ or ‘people who were bishops’ or ‘people who were kings’. We should be capable of extending this same presumption to the less fortunate.

Nevertheless, it is understandable if proponents of the term ‘enslaved people’ are concerned that the connotations of degradation which ‘slave’ has long had will get in the way of their audience's sympathy. More so than ‘conscript’ or ‘convict’, the term ‘slave’ has had a strong association with livestock.⁶² For example, in the Salic Law of the Franks, slaves were ranked with livestock in clauses dealing with the theft of property.⁶³ Indeed, there was disquiet in late antiquity against slaveholders categorising people as livestock through calling them ‘slaves’; historians today are not the first to have identified that the term ‘slave’ is problematic. Gregory of Nyssa objected to the boast of masters that they had got slaves on the grounds it implied ownership over other human beings.⁶⁴ That the term ‘slave’ denied someone's humanity, as if they were naturally property or livestock, is the strongest, and oldest, argument against it.

However, ‘slave’ did not imply that someone was less than human to all early medieval authors.⁶⁵ While Gregory of Tours probably owned slaves himself, a combination of his religious beliefs and the enslavement of his own relatives by sixth-century warlords seems to have led him to look with favour on certain slaves.⁶⁶ He thought that the life of a slave named Portianus, who became an abbot, illustrated the biblical teaching that God would elevate the poor and servile both in this life and the next.⁶⁷ He stated that God had placed Portianus among the angelic choir from which covetous worldly princes were excluded. In other words, through talking about Portianus as a slave, Gregory categorised him not just as a soul-endowed being worthy of God's mercy, but as someone who could become more than human. So even a word like ‘slave’ can have a variety of implications – implications which can change over time. Indeed, most historians who have used the term ‘slave’ over the past few decades would deny that they were implying that slaves were less than human. Rather, they have labelled people ‘slaves’ trusting that most of their readers would conclude that it signifies that a person was treated like property, not that the person naturally was property.

Crucially, the embrace of the term ‘enslaved people’ has followed rather than stimulated the historiographical shift towards emphasising that slaves were historical actors, who had an inherent humanity and agency which slaveholders continually tried to violently suppress.⁶⁸ Vlassopoulos and other historians may be right that we can go further still, but, as his own work shows, this need not be tied to wholesale reworking of terminology.⁶⁹ An appreciation of this may actually help account for why even some historians who prefer ‘enslaved people’ still use terms like ‘slave-produced’ and ‘slave trade’, even though they incorporate the word ‘slave’ and so could be seen as linguistic echoes of the attitudes of masters.⁷⁰ Indeed, the continued use of these terms, rather than ‘produced by enslaved people’ and ‘trade in enslaved peoples’, implies two important logical concessions. First, that concise and familiar expression is a legitimate top priority for historical analysis. Second, that consistently using ‘people-first’ language is not necessary for a reader to understand the humanity of slaves.

Ultimately, slavery as a subject may be too emotionally charged for any term to be devoid of ethical issues. In a discussion among historians of American slavery on the social media site X (formerly Twitter), Michael J. Simpson tweeted that he received pushback on using ‘enslaved’ from ‘Elder people of color … [who] felt that use of “enslaved” was whitewashing’.⁷¹ It was to this that Laura Rosanne Adderley replied, ‘All words I know to talk about enslaved people of African descent in these Americas prove insufficient, both for the brutality against them, and for their remarkable overcoming … “Enslaved” solves some problems, but may create others.’⁷² Slavery may be particularly vulnerable to what linguist Sharon Henderson Taylor and psychologist Steven Pinker have called the ‘cycle of euphemism’ or ‘euphemism treadmill’: ‘give a concept a new name, and the name becomes colored by the concept; the concept does not become freshened by the name’.⁷³ It is easy to dispute Pinker's contention that words can be ‘unexceptional’, and so, by implication, neutral. But if no term can be neutral, is it possible for any term used in reference to slavery to escape our recoil from the horrors of slaveholding practices? As we can see from Simpson's tweet, people have already criticised ‘enslaved people’ as inadequate even for its original context of slavery in the United States, and I have identified other potential issues with the term. It is unlikely that we will arrive at any labels which are beyond dispute.

The politics of the debate

Of secondary importance to the historical and ethical issues is whether there are any professional gains to be had from using ‘enslaved people’ over ‘slaves’. In the increasingly competitive world of faculty politics, it is understandable if the wish not to appear as an out-of-touch reactionary or be left behind is on the minds of some historians. It remains to be seen whether or not the movement against using ‘slave’ will mirror the largely successful campaign against referring to the ‘Third World’, or will start to face substantial pushback, like the movement against using the term ‘Anglo-Saxon’.⁷⁴ Any historian, especially any medieval historian, who carried on using ‘slave’ would be alone neither among scholars of their own period nor scholars of slavery generally. But they should be aware that some have implicitly accused users of the term ‘slave’ of complicity with slavery. Emily Wilson, for example, writes that she uses ‘enslaved’ ‘to avoid complicity with this inherently violent and abusive institution’.⁷⁵ Given that people continue to suffer in slavery today, this is a serious accusation. But it seems unlikely that academics can, in practice, help prevent modern slavery by switching to ‘enslaved person’. Moreover, proponents of ‘enslaved people’ for the victims and survivors of the transatlantic slave trade have not demanded that historians of other periods use that term. Foreman acknowledges that the diversity of slavery means that her recommendations should be treated only as suggestions.⁷⁶ The comments of Adderley referred only to the specific experience of enslaved Africans in America.⁷⁷ Therefore, historians of other periods should not feel obliged to adopt the term ‘enslaved people’. Nonetheless, given that ‘enslaved person’ is increasingly being used in early modern slavery scholarship as well as educational resources even for the classical period, the trajectory is momentarily towards ‘enslaved person’ becoming the dominant academic term. It could be thought prudent to accept it while there is no consensus among scholars, instead of becoming a late hold-out against this transition when there are reasonable, well-intended arguments for it.

However, in trying not to be disconnected from other academics, there is the potential to become disconnected from the wider public: a Google Trends search for April 2022 to April 2023 shows that there were far more online searches for ‘slaves’ than ‘enslaved people’.⁷⁸ Moreover, searches for ‘enslaved people’ were most common – though still much less frequent than searches for ‘slaves’ – in the United States. This is not a trivial issue at a time when journal editors are concerned with search engine optimisation as they try to ensure articles are seen as widely as possible and have an international reach. At least ‘enslaved people’ has an advantage over other academic jargon in that it is hardly obscure; there would be little risk of public confusion over whether it refers to slaves. Moreover, academics can hope to shape and inform the general use of terms around slavery through public engagement and social media. Therefore, they should decide how they want to use their influence after consideration of the historical and ethical issues at stake. It would be a self-fulfilling prophecy to use either ‘slave’ or ‘enslaved people’ out of deference to their perceived popularity in public or academic spheres.

Conclusion and recommendations

Despite the commendable intentions behind it, ‘enslaved person’ is not convincing as a universal substitute for ‘slave’. While ‘enslaved people’ emphasises that someone did not choose to be a slave, this very emphasis means it does not fully accommodate those who sold themselves into slavery. The sheer historical diversity of slave experience also precludes ‘enslaved people’ as well as ‘enslaver’ from bringing us any closer to the perspective of the unfree. These terms are also not essential to recognise the humanity of slaves, which is an ethical imperative we should demand and expect of ourselves and our audiences regardless of our period of study or which word we prefer. However, while it is primarily early medieval exceptions that should caution historians against making ‘enslaved people’ the standard term for slavery studies as a discipline, the term ‘enslaver’ has much more fundamental problems. In attempting to bring moral clarity, it loses historical precision, obscuring the historical dynamics of ownership and manumission. Still, proponents of ‘enslaver’ are on safer ground in highlighting the problems with the term ‘master’. ‘Slaveholder’ or ‘slave owner’ are better than both terms.

If further scholarship can strengthen the case for referring to ‘enslaved people’ over ‘slaves’, historians should be open to using it in most clear-cut cases, while adding clarifying comments when referring to the specific exceptions in which ‘enslaved’ is misleading. Indeed, a number of historians have chosen to use ‘enslaved person’ more often than ‘slave’ without eschewing ‘slave’ altogether – in effect making ‘enslaved person’ the default term and ‘slave’ a synonym occasionally used for linguistic variety.⁷⁹ This seems an attractive compromise, for it keeps the emphasis on the personhood of the unfree, while reserving the possibility of using ‘slave’ for nuanced analysis. Yet any form of mixed use concedes that the term ‘slave’ can be used in contexts that are not dehumanising – or, at least, not so dehumanising as to matter more than the need for historical accuracy and elegant writing.

Whatever decision we as historians make on this issue, it is important that we are prepared to defend it, because this problem is far from resolved. Many have not acknowledged the existing debate, which indicates that some historians are not aware it is taking place. It has even gone unacknowledged in some edited books, even though the variation in terminology by contributors to the same work is jarring. Going forward, historians should explain their choices of terminology much more directly than most have done previously. Rather than take the use of either ‘slave’ or ‘enslaved’ for granted, scholars of different periods need to talk to one another, so that we might better navigate the historical strengths and weaknesses of each term, and how the stakes might vary for slavery in the classical, medieval and modern worlds. Instead of dismissing this dispute over what word to use as irrelevant to the substance of our historical analysis of slavery, we can use it as an opportunity to think carefully about the extent to which scholarship of transatlantic slavery should set the lens of interpretation for slavery studies in general. As the paradigm for slavery, the implications of the transatlantic slave system can both inspire and challenge historians working on other periods. But it also risks clouding the historical variety and complexity of forms which slaveholding practices have taken, especially in the premodern period. After all, I have largely based my arguments on evidence from Europe in the early Middle Ages – perhaps there are corroboratory or indeed counter-examples from other times and places. Future discussion, then, will have the added benefit of illuminating historically contingent aspects in practices and conditions of slavery which are obscured by the common use of one term – whether that be ‘slave’ or ‘enslaved person’.