Ancient Associations and Collective Labor Action in Sarah Bond’s Strike

This essay was part of a 2025 SBL Book Review Panel devoted to Strike: Labor, Unions, and Resistance in the Roman Empire by Sarah E Bond. Read the full forum here. 

1. Summary of thesis

With this fascinating new book, Sarah Bond provides a new model for understanding the relationship between ancient associations and labor in Rome, Italy, and the western provinces primarily, but also in the Greek East. The book analyzes the better-known instances of collective labor action in antiquity, of course (e.g., IEph 215 [Ephesus; II CE]; CIL 3.14165.8 [Arles, late II/early III CE]; Acts 19.23-41). However, whereas these episodes have been seen usually as unreflective of the activities and function of occupational associations in antiquity, Bond demonstrates that they were actually mere illustrations of an ongoing, often silenced, broader history of collective bargaining, striking, and other violent and non-violent collective action behind the transformation of Roman society throughout the Republican and Imperial eras.

2. Ancient Associations after Bond’s Strike

I would like to highlight two standout contributions the book makes relating to ancient associations. First, the initial three chapters, which cover the Roman Republic, read like a new history of Rome, one that shows how ingrained collective labor action really was in Roman society. Specifically, associations, broadly construed, are shown to have been important players in the plebeian secessions (494-287 BCE) and slave wars (135-71 BCE). The strategies enacted by early plebeians are described in ways that bring to mind not only Marxist approaches to the struggle of the orders, but also details more familiar from some of the famous aforementioned instances of labor disputes that involved associations: for instance, the secession to the Sacred Mount in 494 BCE is akin to a strike resulting in “[b]usiness and activities in the city [coming] to a halt” (p. 26), which ultimately resulted in “a deal … worked out” (p. 26).

To be sure, Bond does not argue that early plebeians and their allies formed one giant occupational association for the whole of the Republican period. Rather, labor-related agendas and activities in episodes of plebeian (and ally) collective labor action are compared with – and found to be similar to – those of durable (as in, non-episodic) Dionysiac artist and athletic guilds, and a host of other “associations” that are analogous to modern labor unions. Bond emphasizes the shifting nature of group identities in broadly-construed early Roman associations, specifically the associations that advocated for their members’ greater political agency and quality of life. These collectives could consist of less-wealthy plebeians, migrants or foreign workers, Roman workers, and freed-people.

Many of the demographics from which “members” came were in Rome earlier than is sometimes contended on the basis of literary evidence. For example, it is shown that by the 8th century BCE, Greek merchants had a presence in Rome, and that workshops flourished beyond what can be called domestic production (p. 30). Overall, Bond illustrates that a diverse early “working order” made up an important sector of the Roman populace, certainly already by the Early Republic, and that with the help of organizational principles known from the associations, this ordo fought for wealthier plebeians to have access to political agency, and for the rest to attain relief from debt bondage.[1]

By the time of the three servile wars in the Late Republic, new issues such as landlessness and involuntary labor conditions united slaves and free workers in a new era of collective action. The associations loom large here, too. Specifically, while domestic masters and ludi managers were creating involuntary associations (familiae) of household slaves and gladiators for the purpose of efficiency in organizing and deploying their servile workers, these very associations fostered camaraderie and “a sense of identity” among members (p. 66). And in fact, all three slave wars began with these familiae (pp. 62-64). Moreover, in the third war, Spartacus recruited from workshops (ergastula) in locales surrounding Capua (p. 66). So, as during the conflict of the orders when the plebs had allies from various demographics, here too the escaped gladiators from Lentulus’s school seem to have tapped into an underlying workers’ network that cut across different non-elite demographics. The associations (or at least structuring principles also found in associations) helped organize and unify these collectives in all cases: there is evidence that the slave armies in these wars used language, leadership structures, and collective actions – such as commemorating deceased members and other religious activities – that were all characteristic of associations (pp. 66-69). While these armies consisted, in large measure, of members involuntarily enrolled in associations (familiae), Bond illustrates the agency these associations provided them, anyway.

Bond’s compelling history of Republican Rome made me wonder what a parallel study of work stoppages and collective labor action in Bronze to Classical-era Greek history might reveal. Already in Homer, Achilles’ “strike” responds to Agamemnon’s unfair treatment of him, and Agamemnon’s unvirtuous character, specifically his untrustworthiness (ἐπεί μ᾽ ἀφέλεσθέ γε δόντες [Iliad 1.299]) – later, his “imprudence” or “shamelessness” (ἀναίδεια, [Iliad 9.372]). Achilles attempts to generate a broader consensus on these points and so manufacture wider disobedience against Agamemnon (τῷ πάντ᾽ ἀγορευέμεν ὡς ἐπιτέλλω ἀμφαδόν [Iliad 9.369-370]). Later cases of family and clan-based labor action in the tangled history of the origins of the polis and hoplite phalanx actually mirror very closely Bond’s history of the Early Republic. Moreover, stories of negotiations and revolts within Greek leagues and federations – which themselves invite comparison with private associations – beginning with the account of Ionians ramming Pausanias’ ship in protest to Spartan mistreatment of Hellenic League members (Plutarch, Aristides 23.5; cf. Thucydides 1.95.4), leading to the origins of the Delian League (Thucydides 1.95.6-1.96.2) can be documented in Classical and Hellenistic eras. That Bond’s early chapters inspire applications of her research to other historical periods speaks to the intellectual depth of her treatment of the Roman Republic.

A second of the book’s contributions to the study of ancient associations relates to the value of thinking about collegia broadly to some end rather than finding minor differences that justify less ambitious comparative projects.[2] The book brings together episodes involving Jewish groups, Christian assemblies, involuntary slave associations, plebeian collective action, gangs, the praetorian guard, and assemblies of people in taverns and bars in its analyses of labor action. Neighborhoods and associations, moreover, are often functionally and demographically synonymous in this book.[3] In defense of this approach, Bond underscores the following reality during the principate: “Whether they self-professed as collegia, factiones, corpora, thiasoi, or any of the litany of terms for collectives, they were often treated similarly by Roman authorities during the institution of a ban emanating from the emperor, senate, or governor – whether or not the collective thought of itself reflexively as an association or a political club” (p. 109).

In support of Bond’s inclusive approach to studying associations, it might be added that modern attempts to assert control over which groups “count” as associations can be traced back to a strategy used by Roman authorities to further their own goal of controlling which associations counted as legal. The ancient criteria include concepts and categories such as “ancient,” “useful,” tenuiores, religio, and “political.” These criteria are all, to quote Andreas Bendlin on religio, “underdetermined.”[4] They can be deployed imprecisely by authorities against collectives they wish to crush or prevent from forming. For example, when Trajan wanted to deny Nicomedia a “fire brigade” of fabri, he obfuscated by claiming that they may turn into a hetaeria (Pliny, Letters 10.34; cf. Bond pp. 122-123). In fact, the word hetaireia/hetaeria might at once be used for a political association (see Plutarch, Aristides 2.4; Pliny, Letters 10.96.7) and at another time be the Greek equivalent of the Latin collegium (Gaius, Digest 47.22.4). Sub-categorization is the rhetorical game played by Augustus and Trajan (and others) to suit their own interests. Clodius, on the other hand, saw what was happening on the ground, and so he didn’t pull supporters from a single sub-category but rather noticed how the associations cut through, and gave collective organization to artisans and other workers in shops and workshops, slaves and freed persons, plebs, and neighborhoods. He was inspired by earlier leaders who saw the same overlap on the ground, such as Spartacus, Caesar, and Cataline.

3. Foundations for New Research

Finally, I’d like to offer a few ideas on new research questions this book raised for me. The first relates to the different Romans responses to (the threat of) labor action. Bond identifies several types of responses, including: brute force against riots and revolts; the creation of underdetermined legal concepts of unlawful assemblies and associations; domestication or co-option of the associations with most leverage, such as those involved in supply chains; deployment of associations of spies; and the broader integration of associations into state bureaucracy in late antiquity.

Two additional Roman responses might be union-busting and state-sponsored disinformation campaigns. Building on the work of Jonathan Perry,[5] it might be said that the proconsular edict from Ephesus in response to the marketplace bakers’ undefined collective action (IEph 215) attests to attempted union-busting. Interestingly, the proconsul did not ban the associations of bakers in Ephesus, nor punish any individual members, nor specify any penalties for future labor action. Rather, he threatened that future “seditious” assemblies were prohibited (e.g., συνέρχεσθαι τοὺς ἀρτοκόπους κατ’ ἑταιρίαν [line 8]; συνιὼν παρὰ τὰ διηγορευμένα [line 10]). Buckler wasn’t quite sure what to do with this. He proposed, in short, that strikes happened rarely and as long as they didn’t create disorder, the authorities let them occur, as here, without much in the way of punishment (Buckler 1923, 45).[6] Bond’s book renders this explanation insufficient. Perry, though, notices the decree’s emphasis on political speech, which was presumably spoken by union bosses. This can be seen in the restored word, σύλλογον, in lines 2-3 for “reckless, evil speaking,” as well as terms such as ταραχή (“disorder” [line 2]) and θόρυβος (“uproar” [lines 2, 10]) that convey, in Perry’s words, “the rumble of the crowd, inflamed by the speaker, and preparing for action as a result of what they have heard” (2015, 199). Perry’s interpretation of some of these peculiar features of the edict is that the unnamed governor was trying to “detach the average collegiatus from the organizational élite. The proconsul is actually targeting these dangerous speakers, and he wants to make sure that the other bakers recognize that he is on their side” (2015, 200). In other words, this is attempted union-busting.

The fragmentary proconsular edict from Pergamon, dated to the second century, shares some of the same peculiarities noticed above, and may also represent a union-busting strategy.[7] The specific labor disruption in the Pergamon case is a delay in completing a structure that artisans and others were contracted to finish. The fact that the proconsul of Asia became involved attests to the importance of the project, and it is suspected by Buckler and others that some sort of labor action may have been behind this particular delay (1923, 34). As in Ephesus, there appears to be a proconsular attempt to exploit factions or dissent within the relevant labor associations as a way to neutralize political speech. Specifically, the artisans who elect to come to work on site will not be punished (lines 4-5), yet others who refuse to show up (and so follow “union” directives) will be fined (lines 9-10).

This form of state intervention (i.e., attempted union breaking) raises the question of whether rank and file members of occupational associations were ever truly at risk of expulsion. The tactic of expelling only some practitioners of a purportedly problematic profession or identity is a known Roman response to various self-authorized religious specialists, such as astrologers and experts in Jewish traditions in Rome, as Pauline Ripat and Heidi Wendt have shown (e.g., Ripat 2011, 121-124; Wendt 2015, 97-126). [8] So, perhaps it was mainly the “political speakers,” ancient equivalents to “union bosses,” who would be vulnerable to expulsion orders. In the future, it would be productive to put the politics of “union bosses” into comparison with that of freelance religious experts.

The practice of “union breaking” and expelling “union bosses” (whom Perry calls political/economic experts in the associations and regards as leadership) rather than targeting “members” is quite interesting in the context of association efforts to come across as perfectly unified. Associations often self-represented as unanimous or nearly unanimous when making decisions.[9] Moreover, bylaws against disorder, including failure to properly honor magistrates and staged absenteeism when officers were scheduled to be honored, are well known.[10] It appears that there would naturally be dissent even without state actors meddling. This provided civic authorities with an opening. Specifically, disinformation campaigns – or any form of interference that exploited internal dissent in associations – could effectively nullify the leverage and agency of an association or economic network. In Perry’s reading of the proconsular edict from Ephesus, this psychological tactic was part of the response strategy.

 A second way that this book inspires further research relates to the concept of a strike in antiquity. Here, I’m drawing inspiration from Elizabeth Carney’s article on mutiny in antiquity (1996).[11] As Carney shows, mutiny has tended to be understood as a lack of discipline – specifically, a failure to obey orders. Yet, what is mutiny in a culture where discipline isn’t understood to be based on obedience? As Carney shows, Greek and Macedonian cultures viewed mindless discipline as cowardly, not orderly. So, in these cultures, the modern concept of obedience as a virtue is anachronistic (1996, 20 and throughout). Back to the Iliad for example, Achilles protests that if he obeyed Agamemnon mindlessly, he’d gain the reputation of a worthless coward (δειλός τε καὶ οὐτιδανὸς καλεοίμην [line 293]). This is why it is rare to find instances of army commanders punishing soldiers for lack of obedience in such cultures (Carney 1996, 22-24). In fact, in Homer, Classical Greece, and (Homeric) Macedonian societies, several virtues undermine the merits of mindless discipline, including: outspokenness (Homeric or democratic), bravery, and competitive excellence. As Carney proceeds, she discovers that in Greek and Macedonian armies, order was maintained by difficult leadership techniques the commander had perfected, not simple orders with the expectation that everyone listen because orderliness and chain of command were inherently virtuous. These techniques are exhibited by the Classical-era Macedonian king, for instance: providing clemency as a way of endearing himself to the soldiers; creating the appearance of fair rewards and punishments; maintaining social relationships with soldiers; organizing leisure activities for soldiers (e.g., hunting, games); and showing a paternal concern for troops so that he is seen as someone worth listening to.

Overall, order was based on a proper relationship between king-commander and his troops (Carney 1996, 31 and elsewhere). With this, Carney has provided a good model for understanding order and unrest in military settings. Bond’s analysis on economic matters exhibits points of convergence, and so I wonder if Carney’s work can now be applied more directly to economic matters. Specifically, what unfolds in Bond’s book is a strike concept that is different from earlier efforts to define strike activity; in Bond’s book, the strike concept is inclusive of “labor-related protests” (p. 1) but without limits pertaining to the involvement of “employers,” or “organized refusals to work,” or over-categorizations of various sub-city level groups.[12] And it seems to me (thanks to Bond) that the same values that Carney found so ingrained in some Greek cultures – outspokenness, bravery, competitive excellence – had their place throughout Roman history, as well. These virtues were even fostered by associations there, and led to a democratic element in Roman society that can help explain Rome’s history of economic activism and collective labor action.[13] We are now far from the assessment that slave labor, integration of associations into state bureaucracies, and “low standard of industry” drained Roman workers from any motivation to coordinate organized labor resistance. I thank Sarah for this important and thoughtful book.

Richard Last is Associate Professor, Ancient Greek and Roman Studies Program, Trent University (Peterborough, Canada).

[1] It should be highlighted that Bond is careful to acknowledge that our later records of the early republican history are “quasi-historical” (p.35), “myths” (p.34), and “lore” (p.43).

[2] For the open-ended definitions of associations which is based on group designations and practices, see pp.10-11.

[3] See, for instance, Bond’s important point that “[n]eighborhoods, associations, and collegial leadership were firmly linked” (p.74) in the context of the ban of collegia and ludi Compitalicii in 64 BCE (Asconius, Commentary on Cicero, Against Piso 320.8-9).

[4] Andreas Bendlin, “Associations, Funerals, Sociality, and Roman Law: The Collegium of Diana and Antinous in Lanuvium (CIL 14.2112) Reconsidered,” in Markus Öhler, ed., Aposteldekret und antikes Vereinswesen: Gemeinschaft und ihre Ordnung (Tübingen: Mohr Siebeck, 2011) 207–296, at 225.

[5] Specifically, Jonathan S. Perry, “’L’État intervint peu à peu’: State Intervention in the Ephesian ‘Bakers’ Strike,’” in Vincent Gabrielsen and Christian A. Thomsen, eds., Private Associations and the Public Sphere: Proceedings of a Symposium held at the Royal Danish Academy of Sciences and Letters, 9-11 September 2010, Scientia Danica Series H Humanistic 8.9 (Copenhagen: The Royal Danish Academy of Sciences and Letters, 2015) 183-205.

[6] William H. Buckler, “Labour Disputes in the Province of Asia,” in William H. Buckler and William M. Calder, eds., Anatolian Studies Presented to Sir William Mitchell Ramsay (Manchester: Manchester University Press, 1923) 27-50.

[7] MDAI(A) 24 (1899) 197 no. 62 = IGR 4.444 (reign of Hadrian)

[8] Pauline Ripat, “Expelling Misconceptions: Astrologers at Rome,” Classical Philology 106 (2011): 115-154; Heidi Wendt, “Iudaica Romana: A Rereading of Judean Expulsions from Rome,” Journal of Ancient Judaism 6 (2015): 97-126.

[9] For example, vote counts are recorded so as to assert unity and disguise dissent. When there is a unanimous decision, this is remarked upon; when there is dissent, only the majority decision is recorded. In CIL 14.2112, the unanimous vote ([placu]it universis ut …) in column 1, line 20 can be compared with all the other items voted upon, which seem to have been approved by majorities but not unanimously (e.g., placuit ut…[column 1, line 21]).

[10] For the offense of neglecting one’s duty of properly honuring an association magistrate, see IG 2(2) 1292, lines 15-20 = GRA 1.15 (Athens; 215/214 BCE).

[11] Elizabeth Carney, “Macedonians and Mutiny: Discipline and Indiscipline in the Army of Philip and Alexander,” Classical Philology 91 (1996): 19-44. For discussion of mutiny in Bond’s book, see e.g., pp. 26-27, 70-71; 137.

[12] Striking (as with mutiny) has been defined so narrowly that, until Bond’s book, we have missed how ingrained public debate and activism was in Roman society (at least when the topic of labor disruptions arises). It has been missed because Buckler limited strikes to “artisans… and … their conflicts with employers” (“Labour Disputes,” 27). Likewise, Ramsay MacMullen rejected “a refusal to work” as too broad a definition of striking activity (“A Note on Roman Strikes,” Classical Journal 58.6 [March 1963]: 269-271, at 269). Instead, MacMullen opted for “an organized refusal to work” and indicates that this would eliminate from consideration many of the examples that are in fact covered in Bond’s book (MacMullen, “Roman Strikes,” 269), such as the plebeian secessions (covered by Bond on pp.21-43), the anachoreseis [ἀναχώρήσεις] of Egypt (covered by Bond on pp.110-112), and labour disputes in late antiquity (covered by Bond on pp.132-180). Later, MacMullen agreed with the majority opinion that there is “close to no evidence at all for strikes” (Roman Social Relations, 50 B.C. to A.D. 284 [New Haven, Conn.: Yale University Press, 75), and here seems to define strike activity as “associations seek[ing] benefits [connected] with their work” (MacMullen, Social Relations, 75). More recently, Koenraad Verboven limited strikes to cases of “labour disputes” in which “guilds seem to have put pressure on public authorities” (“Guilds and the Organisation of Urban Populations during the Principate,” in Koenraad Verboven and Christian Laes, eds., Work, Labour, and Professions in the Roman World [Leiden: Brill, 2017] 173-202, at 194). These definitions tend to narrow the scope of investigation in different ways (e.g., identity of the strikers is essentialized as “artisans;” the underdetermined criterion of “organized” must be met by striking collectives; and “guilds” traditionally understood must be present).

[13] See, for instance, Tacitus, Annals 1.16-30 on the legionary mutinies in Pannonia and Germania just after Augustus’ death in 14 CE.