Donaldson IML (2016). Reflections on translating passages on ‘empirical’ and ‘dogmatic’ medicine in Celsus’s De medicina.

© IML Donaldson, University of Edinburgh & Royal College of Physicians of Edinburgh. E-mail:

Cite as: Donaldson IML (2016). Reflections on translating passages on ‘empirical’ and ‘dogmatic’ medicine in Celsus’s De medicina. JLL Bulletin: Commentaries on the history of treatment evaluation (

‘It has been my principal care to convey the precise meaning of my author, and also to preserve the genius of his stile, where the English idiom would allow. I have likewise been careful not to wrest any expression of Celsus, in order to deceive the reader into a greater opinion of his knowledge, than he really deserves. His merit is sufficiently great without pretending to find in him any discoveries, the honour of which is due to the moderns. Every man of learning, who is acquainted with the state of physic among the antients, and knows how far it differs from the modern, must be sensible of the difficulty of translating an author so elegant and concise, with the strictness necessary in a work of this nature. Such judges, I hope, will censure the faults, which cannot escape their observation, with the candor inseparable from true criticism.’

Quotation from what seems to be the first English translation of Celsus. Greive J 1756. Preface xvii-xviii.

In this commentary I reflect on questions that arose when translating passages about ‘empirical’ and ‘dogmatic’ medicine in Celsus’s de Medicina for the James Lind Library (Celsus 1478), and in Howick’s preparation of an article about these (Howick 2016).

All ancient texts pose problems for the translator; technical texts – including those on medicine – are particularly difficult because one is often uncertain of the precise meaning of the technical vocabulary. At first sight, we might expect Celsus to be easier than later Latin texts. Celsus wrote during what was still the golden age of classical Latin. His literary style attracted the praise of his contemporaries – it is in many ways Ciceronian – so we are not faced with the problems of mediaeval texts for which there is no single satisfactory dictionary of the vocabulary, or of renaissance works for which we often have to rely on such bi-lingual dictionaries as are available (Estienne’s great Dictionarium Latinogallicum from the 16th century is probably still the most useful). For the vocabulary of Celsus, the incomparable Latin-English dictionary of Lewis & Short, (L&S), now well into its second century without revision, should surely suffice – and it does.

But Celsus does present special problems. I shall come in a moment to the comment by Greive cited above. But there is an anterior problem – one that is not unique to Celsus but is perhaps particularly relevant to him – the establishment of the Latin text. The text now generally used is due to Marx in 1915; it is a scholarly recension of the available manuscripts and is probably the best we are ever likely to have. Obviously, this was not the text used by Greive for his translation, though it is used by the Loeb edition. The establishment of a scholarly text which notes variants is, of course, a common step in the understanding of most ancient writings. Why should Celsus be any different from other authors? In the wide sense, perhaps he is not. But, because of the relatively late re-discovery of manuscripts of his text, we do have some knowledge of how the foundations of our current text were built. The first scholarly recensed text is fairly certainly that used for the editio princeps, the first printed edition produced in Florence in 1478. It was constructed by Bartolomeo Fonzio (Bartolomaeus Fontius). Christies’ catalogue note for the sale of the Norman copy of De medicina in March 1998 summarises most of what we know of Fonzio’s sources:

Almost unknown in the Middle Ages, De medicina was rediscovered in the fifteenth century when several early manuscripts came to light. The first edition was prepared by the humanist Bartolomeo Fonzio using a codex copied by his brother Niccol from a now-lost source and incorporating corrections based on a ninth-century copy of the text (now Florence, Laur. 73.1). This manuscript may have belonged to Francesco Sassetti, the wealthy Medici banker and collector, whom Fonzio thanked for helping him obtain access to manuscripts (A.C. de la Mare, ‘The Library of Francesco Sassetti’, in Cultural Aspects of the Italian Renaissance, ed. C.H. Clough, Manchester 1976, p.170 and no.78).

What Christies’ note does not tell us is what Fonzio himself said about his establishment of his text. Here he is addressing his patron, Francesco Sassetti, and referring to the use of material in Sassetti’s collection:

…. Ab his ego: ut multorum utilitati consulerem ; Cornelii Celsi scriptoris gravissimi atque eloquentissimi de medicina libros imprimendos curaui. In quibus tanta fide & diligentia usus sum ut manes Cornelianos nobis gratiam habituros : cunctos autem medicinae ac litterarum percupidos certe acturos existimem. Nam cum eius libris pluribus essent in locis temporum inuria mutilati atque inversi : uetustis exemplaribus tua opera e gallia conquistis in unum omnia saepius conferes inanti quum ferme statum redegi. Celsum igitur squallidum antea & deformem / renouatum nunc & iam prope in suam faciem restitutum ob eam maxime causam ad re mitto: ut cum particeps mecum in eo corrigendo laboris fueris: in eodem nunc meliore comptiore que hospite relegando partem etiam capias uoluptatis. Vale.

… From these, having considered how useful it would be to many, I have undertaken to have printed the important and elegantly written books of Cornelius Celsus about medicine. In doing this I have been so faithful and diligent that the shade of Cornelius will be grateful to us: I find their presentation altogether excellent both as medicine and as literature. For many of his books were in places damaged and perverted by the passage of time: from ancient examples brought from Gaul by your offices I have made one [text] in consultation with the wisest collaborators [replacing] lacunae with firm text. Therefore, Celsus, who was formerly squalid and deformed, is now renewed and almost restored in appearance by my strenuous efforts: since you also took part in my editorial work may you share in the pleasure of it now much more polished than it was. Farewell.

(Introduction / dedication, Celsus, 1478. Translation IMLD)

Modern commentators suggest that ‘from Gaul’ meant not from France but from Cisalpine Gaul – northern Italy. For me ‘e gallia conquistis’ – ‘looted from Gaul’ – conjures up irresistible images of Julius Caesar’s conquests – and I do wonder if such an analogy may have been in Fonzio’s mind when he wrote it – but of that, no matter here.

The point is that the first printed text was a construction from diverse manuscripts, as is the now ‘standard’ text of Marx translated in the Loeb edition. I summarised a little more information about the editio princeps of 1478 in two short pieces in 2014 (Donaldson 2014a; 2014b).

But some of Fonzio’s material was no longer extant for Marx. I don’t mean to suggest that I think our text of Celsus is unreliable in any major way, but we should just be conscious that what we have is a reconstruction of reconstructions of many-times-copied-by-hand material long removed in time from the original author. And, on the accuracy of scribes, Petrarch, in the fourteenth century when all copying of books depended on scribes, was trenchant – for him, speaking from bitter experience, they were hopelessly inaccurate…..

But, one may object, surely this is true of most classical texts? Indeed it is. So, why does Celsus present any particular difficulty?

The first reason is because Celsus is to some extent technical and the technical vocabulary of classical Latin is both limited in extent and not always easy to interpret in modern medical terms with any certainty. Remember that Celsus predates Galen and that the Latin interpretations of Galen’s Greek texts came to colour medical thought and to influence its technical vocabulary greatly for nearly two millennia. Celsus had no such vocabulary. Greive, in his eighteenth century translation, is very conscious of this. So one needs to be cautious and to avoid over-precise attempts at interpretation and, above all, to avoid wishing on Celsus interpretations that depend on knowledge we know he did not have.

The second difficulty arises, oddly, from one of Celsus’s strengths; as Greive said, Celsus is ‘an author so elegant and concise’. He was admired in his own time for the literary quality of his writing. Classical Latin of the golden age is admirable in its ability, in the right hands, to express a great deal in few words. Where an author speaks of non-technical matters – of human behaviour, of battles, of love, treason, honour, heroism – such concise prose is powerful, partly on account of what one may call the penumbra of meaning carried by many individual words. And, of course, by literary allusion. Mention Troy and which reader does not recall the Iliad? Can one hear of Nestor without thinking of dignified grandeur in old age? Of Odysseus without cunning? Of Hephaestus sweating over the armour for Achilles without recalling his lameness and its cause? Sadly, many modern readers are quite deaf to these allusions. Nevertheless, even the modern reader can perhaps accept the existence and influence of the penumbra without needing to agonise over just what shade of meaning was intended. But, for technical material, the opposite is true; the translator is forced to choose between a number of possibilities and to try to force from the context of the text clues as to which choice to make. This process is obviously fallible and is coloured not only by the translator’s technical knowledge but by his own reactions, prejudices and so on. Translation, as Bismark said famously of politics, is no exact science.

Fortunately, in the passages from what is now regarded as the Proemium to the Books on Medicine, but was, in the editio princeps, simply included as the beginning of Book I, this is less of a problem because the matter is rather general and not very technical. Even so, some choices do have to be made and, in addition, one must decide to what extent to amplify Celsus’s terse words. Some examples appear in my comments below. Unless otherwise noted, the Latin text I quote is that of Marx as given by the Loeb edition.

Quotation 1

Etiam sapientiae studiosos maximos medicos esse, si ratiocinatio hoc faceret: nunc illis verba superesse, deesse medendi scientiam.


…Even philosophers would have become the greatest of medical practitioners, if reasoning from theory could have made them so; as it is, they have words in plenty, and no knowledge of healing at all. …


That even the philosophers must be allowed to be the greatest physicians, if reasoning could make them so; whereas it appears, that they have abundance of words, and very little skill in the art of healing.


In addition, if this reasoning held, it would make doctors of those most steeped in wisdom: while they have a surfeit of words, they know nothing of medical treatment.


…make the greatest doctors of those steeped in wisdom…

The adjective ‘maximos’ could qualify either ‘studiosos’ or ‘medicos’. I think ‘the greatest doctors’ is perhaps the more likely.

Next, all the previous English translators take those ‘studiosos sapientiae’ to be ‘philosophers’.

I think this may reflect, first, that successive translators use phrases so similar to their predecessors that it sometimes seems that they have largely copied them rather than re-considering the Latin. In this particular case, translating studiosos sapientiae (‘students of wisdom’) as ‘philosophers’ would have been quite natural in the 18th century. After all, the Latin is almost a translation of the Greek ‘lover of wisdom’.

But Latin did contain a specific word for philosopher – which was ‘philosopher’! And Celsus did not use it in this passage.

I think Celsus was just referring to what one might call ‘learned men’ and not to those who practised the trade of philosophy. In eighteenth-century English, ‘philosopher’ was widely used as a general term for ‘learned man’ – often in distinction to a ‘plain’ or ‘practical’ man – but this usage has now largely died out in vulgar English. So, now, ‘philosophers’ is generally limited to those who practise that academic specialty. Thus using it in a contemporary article might perhaps mislead the modern reader?

This is perhaps rather trivial. But, to me, ‘philosophers’ seems wrong – for the modern reader – because it seems to make Celsus single out what, nowadays, is a sub-class of the educated, and I do not believe that Celsus was doing that. Rather, his language suggests to me the sense that, if just having a ‘rational’ argument sufficed, any educated man – any Thomas, Ricardus or Henricus, if you will – would, simply by virtue of the force of that argument, be made a medical practitioner. That is why my translation says that the reasoning would make the wise men doctors.

Quotation 1 second part

The next part of the quotation is trickier…..

Cum igitur illa incerta, inconprehensibilis sit, a certis potius et exploratis petendum esse praesidium, id est is, quae experientia in ipsis curationibus docuerit, sicut in ceteris omnibus artibus.


Since, therefore, the cause is as uncertain as it is incomprehensible, protection is to be sought rather from the ascertained and explored, as in all the rest of the Arts, that is, from what experience has taught in the actual course of treatment


Seeing then this last is uncertain and incomprehensible, ’tis much better to seek relief from things certain and tried; that is, from such remedies as experience in the method of curing has taught us, as is done in all other arts

IMLD – literal translation – first attempt.

Therefore, since that is uncertain and incomprehensible, help should be sought preferably from the reliable and tested; that is, [from] that which experience has taught in these treatments, as is the case in all the other Arts.

All versions are defensible; all the shades of meaning – and some more (!) – are implicit in the Latin, but I think ‘protection is to be sought’ as Spencer has it, is not very helpful to the reader. I think Celsus means ‘trust that which has been proved by experience’. He goes on to talk about steersmen learning from experience.

Incidentally, in the sixteenth century, Jacques Dubois (Iacobus Sylvius), in his vituperative attack on Vesalius’s use of illustrations in his Fabrica, quotes Galen as making a rather similar remark (I have not succeeded in finding the original) about learning from practice and not from looking at pictures. The problem is the meaning of ‘praesidium’.

Lewis & Short begin their entry:

‘a presiding over; hence defence, protection, help, aid, assistance; esp. of soldiers who are to serve as a guard, garrison, escort or convoy.’ Then follow many examples, all military. But a more general use as: aid, help, assistance of any kind is also recorded and finally: ‘In particular a remedy against diseases.’

For this meaning, the authority is Pliny – a contemporary of Celsus – and the examples quoted are: ‘aurium morbis praesidium est’ [it is a remedy against disease of the ears] and ‘contra serpentes praesidio est’ [‘a remedy against [bites of] serpents’]. I have checked Pliny’s text; L&S’s interpretation seems to me the only one reasonably possible – that, here, ‘praesidium’ means remedy or cure.

So, perhaps Celsus was really not saying anything about seeking ‘help’ or ‘aid’ or ‘protection’ but rather simply that:

‘a remedy should be sought from the reliable and tested; that which experience has taught in these treatments.’ This led me to:

IMLD – second attempt:

Therefore, since that is uncertain and incomprehensible, remedies should be sought preferably from the reliable and tested; that is, [from] that which experience has taught in these treatments, as is the case in all the other Arts.

This has the merit of being entirely clear and not raising any likely irrelevances about protection and aid; but is it what Celsus meant? I am happy to defend my version – but, really, I can’t be certain it is ‘right’.

Arrogantly, I prefer my second translation to all the others.

Quotation 2

This is the most problematic of the marked sections.


They do not deny that experience is also necessary; but they say it is impossible to arrive at what should be done unless through some course of reasoning. ….. Again they say that it makes no matter whether by now most remedies have been well explored already… if, nevertheless, they started from a reasoned theory; and that in fact this has also been done in many instances.

The whole text of this passage in the Loeb edition is:

Neque vero infitiantur experimenta quoque esse necessaria, sed ne ad haec quidem aditum fieri potuisse nisi ab aliqua ratione contendunt: non enim quidlibet antiquiores viros aegris inculcasse, sed cogitasse quid maxime conveniret, et id usu explorasse, ad quod ante coniectura aliqua duxisset. Neque interesse, an nunc iam pleraque explorata sint…* si a consilio tamen coeperunt. Et id quidem in multis ita se habere.

*Marx provides the text below for the lacuna, but does not reveal its source.

“deesse apparat ‘si quotannis tamen nova remedia inveniuntur, neque dicendum esse antiquiores experimentis esse usos . . .’”, Marx (1915) p. 20. Spencer, in the Loeb edition renders the interpolation: ‘if new remedies nevertheless are found every year, nor must we say that the ancients went by experience…’

The main problem with this quotation is that it is not easy to understand, and Spencer’s version, though it ‘translates’ the Latin words, does not make much sense.

The Loeb edition (based on Marx’s Latin text) indicates that text is missing at the asterisk. Marx supplies text here which he marks deesse apparat (appears to be missing) – see above. Unfortunately, he does not seem to indicate his source for the apparently missing text which, one supposes, was not available to Fonzio in the fifteenth century since, as will be seen from the images (Celsus 1478), the editio princeps does not include the missing text nor mark any lacuna after ‘explorata sint’. The interpolation emphasizes that, here, Celsus is denying that all ancient physicians were necessarily Empirics. The editio princeps has ‘inficiantur’ for ‘infitiantur’, but this is simply a variant without significance.

It would be well to emend the English translation to include Marx’s additional text since it fills an obvious lacuna and makes Celsus’s current point more forcefully as well as making more sense. The last phrase – ‘and that in fact this has also been done in many instances’ – seems to make little sense even after the addition of the text presumed to be missing. Note that, in the Latin, the phrase stands as a new sentence and it would be best to respect this. But I suspect there is still something wrong with the Latin text.

I suggest, as a translation of Marx’s text – without any claim that its meaning is clear! :


They do not deny that experience is also necessary; but they say it is impossible to arrive at what should be done unless through some course of reasoning. …

They say that it makes no matter whether by now most remedies have been well explored already. If, however, new remedies are found every year; we must not say that the ancients went by experience if, nevertheless, they started from a reasoned theory; And that in fact this has also been done in many instances.

Grieve’s eighteenth-century translation is considerably different and is worth examining. Here is his version of the whole passage:


Nor do they deny experience to be necessary, but affirm, it cannot be obtained without some theory ; for that the more antient practitioners did not prescribe any thing, at hazard, for the sick, but considered what was most suitable, and examined that by experience, to which they had before been led by some conjecture. That it is of no moment in this argument whether most remedies were discovered by experiment, provided they were at first applied with some rational view: and that this holds in many cases;

This version has the merit of making a good deal more sense than Spencer’s. But, how compatible is it with the Latin? Greive used an 18th century recension of the Latin text by Almeloveen. Examining this we find ‘an initio pleraque explorata sint’ in place of ‘an nunc iam pleraque explorata sint’. It is from this that Greive derives ‘provided they were at first applied with some rational view’ rather than Spencer’s ‘by now well explored’ – a completely different statement. Greive’s presentation of the argument that Celsus is setting out, about the beliefs of the Dogmatics, seems convincing but, I fear, it relies on a great deal of expansion for which there is little basis in the Latin.

The comparison of Spencer and Greive is quite instructive. Spencer sticks rather closely to the source text, ignores the lacuna (though it is marked as apparent in his source text), and the result is not very coherent. Greive, working from a very slightly, but very significantly, different text – but with the unindicated lacuna – seems to use the source text as a rough framework on which he builds a coherent account of what he takes to be Celsus’s account of the belief of the Dogmatics.

Let us now insert Marx’s missing text into Almeloveen’s version (the one Greive used) and see what we can make of the result.

Neque vero infitiantur experimenta quoque esse necessaria, sed ne ad haec quidem aditum fieri potuisse nisi ab aliqua ratione contendunt: non enim quidlibet antiquiores viros aegris inculcasse, sed cogitasse quid maxime conveniret, et id usu explorasse, ad quod ante coniectura aliqua duxisset ducti essent. Neque interesse, an nunc iam initio pleraque explorata sint si quotannis tamen nova remedia inveniuntur, neque dicendum esse antiquiores experimentis esse usos si a consilio tamen coeperunt. Et id quidem in multis ita se habere.

Marx’s proposed interpolation in bold italic; Almeloveen’s variants underlined and the words he omits struck out.

IMLD attempt at translating this:

They do not deny that experience is also necessary; but they say it is impossible to arrive at what should be done unless through some course of reasoning: for the ancients did not treat the sick at random, but, after consideration of what would be most appropriate, tried out that method to which some theory had led them previously. Nor did they consider it important if many [such] things had first been explored even when [if] new remedies are discovered, nor [did they] claim that the ancients based their actions [only] on experience. And certainly matters have often been thus.

This turns out not too dissimilar to Greive’s translation. Of course, what we have done is to take Marx’s edited text – his estimate of the ‘best’ text as composed from the sources available to him – and emend it by injection of the phrase that he believed to have been omitted (one wonders why he did not include it himself) and then reject a few words in favour of those from an earlier editor’s recension. Is this procedure legitimate? In one sense it is what all editors do to make their best text. But, in another sense, it is no such thing. Learned recensions are generally produced by professional palaeographers and classicists selecting from a variety of sources on the basis of enormous experience of similar texts and using various criteria – well-defined or not. I make no pretension to the expertise of such editors. What I have done is much more crude; I have had a single motive – to try to arrive at a version of which I can make some sense – and I have only examined the Latin texts used by Spencer for the Loeb edition of 1935 and by Greive for his edition of 1756. My only excuse for this cherry-picking is that it has produced a text for the passage which seems to support a translation that I believe makes better sense than Spencer made of the passage, but without quite as much invention as Greive used. May the shade of Celsus forgive me if I have traduced his meaning.

Quotation 3

Tum requirunt etiam, quare venae nostrae modo summittant se, modo attollant; quae ratio somni, quae vigiliae sit; sine quorum notitia neminem putant vel occurrere vel mederi morbis inter haec nascentibus posse.


Moreover, they also inquire why our blood-vessels now subside, now swell up; what is the explanation of sleep and wakefulness: for without knowledge of these they hold that no one can encounter or remedy the diseases which spring up in connection with them


They also inquire how it happens, that our arteries rise and fall, from what causes proceed sleep and watching; without the knowledge of which, they conceive it impossible for any person either to oppose the beginnings of diseases, that depend on these particulars, or cure them when formed.

This time, Almeloveen’s text, the one used by Greive, does not differ from Marx except in being punctuated differently and in having ‘submittant’ for ‘summittant’, which mean the same. Spencer and Greive are translating the same Latin text.

Greive, in a footnote elsewhere, remarks on Celsus’s use of ‘venae’ for all blood vessels, saying that, in some places, he clearly means arteries. This seems likely to be correct. In most (but not all) of this passage, Greive both stays closer to the Latin and better expresses its content than the later translators. His ‘watching’ is equivalent to ‘wakefulness’ – an 18th century idiom now largely lost.

The first problem is how to translate ‘occurrere’. Spencer’s ‘encounter’ is unhelpful. Greive’s ‘oppose’ is much better – but is that what Celsus meant? Greive then adds ‘beginnings of diseases’ which is not explicit in the Latin but which, presumably, he somehow extracts from the text. ‘Occurrere’ also has a sense of ‘coming across’ or ‘discovering’, and ‘nascentibus’ of ‘beginning’, though in its context in this passage it is more ‘arising’ rather than ‘beginning’- see below. The penumbra is quite large here.

At first sight, if we were to change ‘watching’ to ‘wakefulness’ and ‘oppose’ to ‘recognise’ in Greive’s version it would make an excellent translation – for me at least. But deeper analysis reveals more problems.

Once again, careful reading of L&S is useful:

For ‘occurrere’, in addition to all the meanings used by previous translators, we find: To obviate, or seek to obviate, to meet, resist, oppose, counteract: and then: To cure or attempt to cure, to relieve, remedy.

Turning then to ‘mederi’ – which more obviously means to heal or cure – we find L&S gives a number of examples of uses in this sense, and in no other senses. So we have two verbs acting on the ‘morbis’ – the diseases – one of which is unequivocal in meaning ‘to cure’, the other of which has a range of possible meanings.

Returning to our text, we have:

neminem putant vel occurrere vel mederi morbis inter haec nascentibus posse.

And we notice that Celsus opposes the two verbs ‘…vel occurrere vel mederi…’ in a standard construction: either to (do something) or to (do something else). So, presumably we should not take both verbs to mean ‘cure’. Our predecessors are sensitive to this; all take ‘mederi’ as ‘cure’ or ‘remedy’ – there is really no other choice. That leaves the question of how to interpret ‘occurrere’ – which is where we started. How do we decide? For me, Celsus is emphasising the need for fundamental knowledge of structure and function. I agree with Spencer that the sense is ‘come across’ or ‘meet’ rather than Greive’s ‘oppose’ – because Celsus does not talk about ‘occurrere’ acting on the beginnings of diseases as Greive translates – in my view incorrectly (see below on ‘nascentibus’).

But Spencer’s ‘encounter’ does not help the English reader to understand Celsus’s meaning, so I would stretch the meaning a little and say ‘recognise’. My interpretation is that Celsus regards knowledge of fundamental structure and mechanisms as necessary, both to recognise disorders and then to treat them.

I did not see any problem in the interpretation of ‘nascentibus’, but Greive’s translation raises one. For me, ‘inter haec nascentibus’ is unequivocal, namely, ‘taking their origin among these’ or ‘arising from these’ where ‘these’ are the causes of the disturbances of the vessels or of sleep or wakefulness. Greive has transposed ‘nascentibus’ to the diseases and translated it as their beginnings. I don’t think the grammar will allow this. ‘Morbis’ is dative because ‘mederi’ usually ‘takes’ its object in that case, and ‘nascentibus’ agrees with it. But ‘inter’ requires the accusative, so its object must be ‘haec’. Greive seems to be taking ‘nascentibus’ as the ‘object’ of ‘occurrere’ (in the dative, which the word-form would allow, and ‘occurrere’ can ‘take’, giving ‘cure the beginnings’; but ‘morbis’ – the form is either dative or ablative – really should not be rendered as the genitive ‘morbi’ ‘of diseases’. And ‘when formed’ is just not present in the Latin. In this case, though one can see how he got to the position, I think Greive is at least perverse – or perhaps just plain wrong.

After all this, how shall we translate the passage?

IMLD’s attempt:

‘They also enquire how our blood-vessels sometimes become prominent and sometimes fall away and what are the causes of sleep and waking; without knowledge of these matters no one can recognise or cure diseases arising from these causes.’

I thought about going a little further and rendering ‘attollant’ as ‘become engorged’ rather than ‘become prominent’. But, by doing this, I would risk wishing on Celsus ideas that, in our time, have implications about the circulation of the blood. I decided not to yield to temptation.

Quotation 3 second part

Praeter haec, cum in interioribus partibus et dolores et morborum varia genera nascantur, neminem putant his adhibere posse remedia, qui ipsas ignoret. Ergo necessarium esse incidere corpora mortuorum, eorumque viscera atque intestina scrutari; longeque optime fecisse Herophilum et Erasistratum, qui nocentes homines a regibus ex carcere acceptos vivos inciderint, considerarintque etiamnum spiritu remanente ea, quae natura ante clausisset, eorumque positum, colorem, figuram, magnitudinem, ordinem, duritiem, mollitiem, levorem, contactum, processus deinde singulorum et recessus, et sive quid inseritur alteri, sive quid partem alterius in se recipit:


Moreover, as pains, and also various kinds of diseases, arise in the more internal parts, they hold that no one can apply remedies for these who is ignorant about the parts themselves; hence it becomes necessary to lay open the bodies of the dead and to scrutinize their viscera and intestines. They hold that Herophilus and Erasistratus did this in the best way by far, when they laid open men whilst alive — criminals received out of prison from the kings — and while these were still breathing, observed parts which beforehand nature had concealed, their position, colour, shape, size, arrangement, hardness, softness, smoothness, relation, processes and depressions of each, and whether any part is inserted into or is received into another.


Besides, as pains, and various other disorders, attack the internal parts, they believe no person can apply proper remedies to those parts, which he is ignorant of, and therefore, that it is necessary to dissect dead bodies, and examine their viscera and intestines ; and that Herophilus and Erasistratus had taken far the best method for attaining that knowledge who procured criminals out of prison, by royal permission, and dissecting them alive, contemplated, while they were even breathing, the parts, which nature had before concealed ; considering their position, colour, figure, size, order, hardness, softness, smoothness, and asperity ; also the processes and depressions of each, or what is inserted into, or received by another part.

Other than the length and complexity of Greive’s sentence, which mirrors the Latin text, I have no quarrel with this, though, this time, I think Spencer is more accurate. I observe only that the English translators, perfectly sensibly, take ‘longeque optime’ to mean ‘the best by far’: but it could also mean ‘took the best method, a long time ago’. It probably does not matter much which version we choose. But one should point out that Herophilus and Erasistratus lived some 300 years before Celsus – who may, perhaps, have had access to some of their works now lost. Celsus is, emphatically, not commenting on contemporary practice in his world of the first century AD.

In Conclusion

All translations are suspect, all are imperfect; some traduce their source. Not only is translation not an exact science, it should be a negotiation, as proposed by the late-lamented Umberto Eco. The negotiation required is between the source-text with its conventions and allusions and the language of the translation in which those allusions, as well as the meaning of the words, should be mirrored so that they affect the reader of the translation in the same way as the original text acted upon its readers. The negotiator, of course, is to be the translator. But, at a distance of some two millennia between author and translator, who shall judge whether the negotiation was just?

I hope that this rather rambling discussion has shown why I agree with Greive that Celsus is not easy to translate into English. At the very least it should, I hope, be persuasive that there is no such thing as ‘THE English translation’. If it also helps to make Celsus’s ideas any clearer that will be a bonus.

To download a pdf of this article click here

This James Lind Library article has been republished in two parts in the Journal of the Royal Society of Medicine 2021;110:359-362 and 2021;110:400-405.


Almeloveen TJ (1722). Ab Aur. Corn. Celsi De medicina libri octo. Padua: Josephus Cominus.

Celsus (1478). De medicina. Florence: Nicolaus (Laurentii).

Donaldson IML (2014a). Celsus: De medicina. Florence 1478. Part 1. J R Coll Physicians Edinb 44:252–4.

Donaldson IML (2014b). Celsus: De medicina, Florence 1478. Part 2. J R Coll Physicians Edinb 44: 344–6.

Eco U (2003). Mouse or Rat? : Translation as Negotiation. London: Weidenfeld and Nicolson.

Greive JA (1756). Cornelius Celsus of Medicine. In eight books. Translated, with notes critical and explanatory, by James Greive, M.D. London: Printed for D. Wilson and T. Durham.

Howick J (2016). Aulus Cornelius Celsus and ‘empirical’ and ‘dogmatic’ medicine. JLL Bulletin: Commentaries on the history of treatment evaluation (

Lewis CT, Short C (1879). A Latin dictionary: founded on Andrews’ edition of Freund’s Latin dictionary / revised, enlarged, and in great part rewritten by Charlton T. Lewis and Charles Short. Oxford: Clarendon Press.

Marx F (1915). Corpus Medicorum Latinorum consilio et auctoritate Instituti Puschmanniani Lipsiensis. Vol. I. A. Cornelii Celsi quae supersunt. Recensuit Fredericus Marx. Leipzig & Berlin: BG Teubner.

Spencer WG (1935-38). De medicina / Celsus; with an English translation by WG Spencer. The Loeb classical library. London & Cambridge, Mass.: Harvard University Press.