At the very outset of his seminal publication L'Enfant et la vie familiale sous l'ancien régime, first published in Paris in 1960 and translated into English two years later as Centuries of Childhood, Philippe Ariès introduces himself to his reader as a ‘demographic historian’.Footnote 1 It is surely one of the most remarkable aspects of this remarkable book that this ‘demographic historian’ proceeds to underpin considerable portions of his arguments about children, education and the family with visual material. Indeed, Ariès not only uses canonical works of art in the thesis he unfolds; he incorporates a wide range of objects, from paintings through to tapestries, from engravings to fans, employing an eclectic approach that foreshadows the more recent emphasis of many art historians on a broadly conceived ‘visual culture’.Footnote 2
In his recent biographical study of Ariès, Patrick Hutton elucidated the critical role of the historian's wife, Primerose, in his turn towards the image. Primerose was trained as an art historian, and it was in her company that Ariès began to rummage in the Salle des Estampes at the Bibliotheque Nationale, in search of images of costume across the centuries.Footnote 3 That project gave way to his ground-breaking study of childhood, and he built up a large repository of visual material, adding works of art viewed in Parisian exhibitions of the 1950s and others found in canonical texts, such as Walther Bernt's volumes on seventeenth-century Dutch art, to the engravings by the likes of Claudine Bouzonnet-Stella that he had collected at the library.Footnote 4 Ariès’s extensive study of education, which forms the central and most substantial part of Centuries of Childhood, makes only sporadic reference to such images, but his much debated discussion of ‘the discovery of childhood’, as well as his analysis of the family in the final part of the book, has this visual material at its core.
Ariès uses these objects in a variety of ways. On a few occasions, he deploys the visual to create a mood. Discussing, for example, a new ‘vague but definite correlation between early adolescence and the typical soldier’ in the eighteenth century, Ariès evokes ‘the handsome young soldier depicted by Watteau’.Footnote 5 He also, and more frequently, refers to the inclusion of certain objects or activities in images; using a picture to illustrate a significant development in children's clothing, for example, or use of a form of discipline such as the birch.Footnote 6 In these cases, the historical artefacts have primary significance, over and above the images in which they have been pictured. Thirdly, and most significantly, Ariès engages with questions of representation. He enquires into the initial absence and subsequent development of pictures of children – particularly of children unaccompanied by adults – and asks whether images of infants across the centuries depict certain distinctively childlike physical and behavioural characteristics. The conclusions he draws underpin some of the key propositions of Centuries of Childhood: the absence of a concept of childhood in the Middle Ages; the emergence of our ‘modern’ understanding of the condition in the 1600s; the free mingling of children in the adult world, before the closeting away of the family in the eighteenth century.
Centuries of Childhood is now over 50 years old, but it holds its place as a pioneering study of childhood, consistently a standard point of reference and orientation. Furthermore, Ariès’s treatment of images as a source material continues to attract comment, both within discussions of the history of children and the family, and within broader considerations of the use of images as historical evidence.Footnote 7 In the following historiographical and methodological discussion, I want to build on a valuable article published in Continuity and Change in 1989 by Anthony Burton, which considered Ariès’s treatment of pictorial and material sources.Footnote 8 Drawing on some of Burton's insights, I will explore the more recent afterlife of Ariès’s approach to visual culture, and its place in histories of children and childhood, concentrating here on paintings. My perspective is, principally, that of an academic art historian. In the next section of this article, I will reflect back on the ways in which Ariès used pictures as historical evidence in Centuries of Childhood, together with the subsequent responses of his critics. I will then move on to consider how his narrative has endured, despite critiques, but also how his key period of change has been quietly, but firmly relocated from the seventeenth to the eighteenth centuries. In particular, I will explore how it has become axiomatic that British child portraiture of this later period shows the advent of ‘The New Child’, and discuss the ways in which an image such as Joshua Reynolds's portrait of John Crewe (c. 1775; Figure 5) might aid our understanding of the histories of children and childhood.Footnote 9
Before going any further, however, it is important to note a persistent problem for anyone considering Ariès’s use of images. Quite simply, most available copies of Centuries of Childhood contain no illustrations at all. Only the first and second French editions include pictures, and the 1962 translation by Robert Baldick is the only illustrated English edition.Footnote 10 Furthermore, the reproductions in the Baldick version amount to a rather meagre 26 grainy plates, and the relationship between those images and the text is loose, so that one has to work hard to link the pictures to specific moments in Ariès’s discussion. The absence of illustrations in all subsequent editions is not merely a source of frustration; it has also significantly affected our understanding of Ariès’s use of visual material.Footnote 11 Practically, it allows basic errors commonly to pass unnoticed, such as his mis-description of the Victoria and Albert Museum's Holme Family (1628) as ‘a triptych showing a little boy and a little girl on the centre volet’ (the children are included in the wings).Footnote 12 More broadly, and more importantly, it has amplified the sense of sketchiness that pervades Ariès’s consideration of the artistic treatment of children across some eight hundred or so years of Western European history. It is considerably easier to make broad claims about the representation of the Infant Christ from the medieval period through to the Renaissance, for example, if those claims are not linked to and grounded in clearly identified and reproduced pictures.
1. ARIÈS’S STORY OF CHILDHOOD AND ITS CRITICS
The most controversial argument based on images in Centuries of Childhood is undoubtedly Ariès’s proposition that no sense of childhood as a distinct condition with distinct needs existed in the Middle Ages, because it was not pictured.Footnote 13 Statements such as: ‘In the tenth century, artists were unable to depict a child except as a man on a smaller scale’, and the conclusion that this was because ‘the men of the tenth and eleventh centuries did not dwell on the image of childhood, and that … image had neither interest nor even reality for them’ became and remain the targets of attack by medieval historians and art historians alike.Footnote 14 Criticisms have, however, come from two distinct perspectives. One approach has been to try and find the ‘missing’ childlike children; to track down medieval images in which infants are not only designated as such by virtue of their smaller scale, but also by particularised behavioural and/or physical characteristics.Footnote 15 The other has been to accept Ariès’s view that children are absent in medieval art, but to dispute its significance. This latter group of scholars has argued that, whilst artists in this period might not have engaged with the distinctive qualities of infancy, we should not extrapolate that society in general was uninterested in those characteristics that distinguish the child from the adult. Just because something does not have a certain form of expression, does not mean, ergo, that it does not exist. On more than one occasion, it has been argued that we are dealing here with a matter of evidence, more than a matter of attitudes. We need to look more broadly at what artists did or did not represent during this period and, in doing so, it becomes apparent that children were not a dominant concern for those creating works of art, largely engaged with religious themes rather than secular subject matter.Footnote 16
After proposing the absence of medieval childhood, Ariès proceeds to trace the gradual advent of more ‘modern’ depictions of infants, including changing images of the young Christ from an archetypal ‘miniature adult’ to representations that engaged more with the ‘graceful, affectionate, naïve aspects of early childhood’.Footnote 17 Ariès is particularly taken with the onset of the artistic tradition in which the infant Jesus is shown naked. He links this in part with a ‘broad surge of interest in childhood’ (presumably as engaging with the distinctive physique of the child, although this is not explicit), and in part with the perceived emergence of an ideal of childhood innocence from a rather raucous past: a past in which both children and adults were happy to make bawdy jokes about young bodies.Footnote 18
This has proved a problematic argument, to say the least. The central place of the child in Christian narrative and doctrine is surely worthy of attention, but the primary response has to be that voiced by Burton: ‘There is an obvious danger in treating depictions of the incarnate Godhead as if they were accurate reports of medieval childhood.’Footnote 19 Burton, amongst others, emphasised that we have to attend to the relationship between these images of Christ and shifting theological doctrine.Footnote 20 A good example of the issues at stake is the way in which Ariès follows a lengthy account of the ‘astonishing liberties’ taken with the young Louis XIII, described by the French Royal physician, Heroard, with a reference to a 1511 Hans Baldung Grien woodcut of Anne apparently trying to tickle Christ's genitalia. Both are presented as evidence of casual familiarity with the body of a child.Footnote 21 However, it is impossible to maintain Ariès’s equation of this story and image since the publication of Leo Steinberg's The Sexuality of Christ in 1984.Footnote 22 Steinberg proposed that an increasing focus on the body of the young Christ in the Renaissance, particularly on his penis, was connected to a new theological emphasis on the Incarnation; on God's assumption of human form. The way in which Ariès deploys the Grien to support a thesis concerning the life of a young French prince becomes clearly unsustainable in light of such analysis.Footnote 23
In the case of these images of Christ, it has been pointed out that one needs a full knowledge of theological doctrine and its relationship with the history of art, before one attempts interpretation. However, most of Ariès’s claims about images of children have more simply attracted the criticism that he lacked sufficient knowledge and understanding of artistic traditions, style and iconography. His proposition that the representation of playful putti in the late fourteenth and fifteenth centuries was part of the inexorable rise of the ‘modern’ child, for example, has been countered by scholars who have emphasised the distinctive role of these little figures in the story of Western art: symbolic, often subversive; participants in an artistic tradition deeply engaged with antiquity.Footnote 24 Perhaps the most notable case, however, is Ariès’s proposal that seventeenth-century Netherlandish painting showed both the routine mingling of children in the outdoor, public world of adults, and the advent of a new emphasis on the family with the child at its centre.Footnote 25 In this context, he reads Jan Steen's Feast of St Nicholas, one of the few specifically noted and easily identifiable images in Centuries of Childhood, as a scene of ‘a quiet family celebration’ (Figure 1).Footnote 26 However, later, seminal work by art historians in this field, spearheaded by Eddy de Jongh, transformed discussions of such pictures. Paintings previously seen as detailed, mimetic transcriptions of the ‘real’ world became complex, symbolically loaded images, and even the most mundane domestic scene became imbued with deep significance.Footnote 27 Erika Langmuir, in her 2006 Imagining Childhood, thus proffered a very different reading of the Steen. She begins by noting the ‘semblance of naturalness’, before homing in on the shadowy male figure at the centre of the picture. She links him with the character of Father Time, proposing that he reminds us ‘of time and its workings: the survival of the Catholic feast from “olden days” and its yearly recurrence; the longevity of family and its perennial regeneration … the ephemeral nature of childhood … The brooding figure may also vouchsafe us uncomfortable glimpses of the future.’ In addition, Langmuir argues that the girl with her doll is ‘surely a future femme fatale’, pointing to the significance of an adolescent boy's treat being concealed in a bed, and proposes that the seated woman in the foreground has more than a whiff of the procuress about her.Footnote 28
Jan Steen, The Feast of St Nicholas, oil on canvas, 82 × 70.5 cm, c. 1665–1668. Rijksmuseum, Amsterdam.
However, one publication on the representation of children in Netherlandish art of this period pushed the new understanding of symbolism to a conclusion that attracted considerable criticism. This was Mary Durantini's The Child in Seventeenth-Century Dutch Painting of 1983, which proposed that such images should, therefore, be seen as having nothing to do with children per se. These pictures, she argued, were vehicles for adult concerns, and we should not see them as telling us anything about either the real lives of children in this period or the ways in which they were perceived; ‘there really is no interest in or comprehension of childhood as such. The children function as commentators for or as embodiments of adult problems, vices or concerns.’Footnote 29 Thus, Durantini comments on the presence of both the ferule and the spinning top (which needs whipping) in Gerrit Dou's Schoolmaster (1645): ‘The necessity of hard work and discipline does not only pertain in childhood … The pressure of the “whip” – be it in the form of self-discipline or some external force – must be applied to adults as well. And it is the adult viewer to whom Dou's schoolmaster is addressing his glance and his message.’Footnote 30
It was broadly agreed that this was taking things too far. Wayne Franits, in a review article, pointed out that the central place of symbolism in seventeenth-century Dutch art should not lead to the conclusion that images of children had nothing to do with childhood. A picture such as the Dou surely relates to concerns about pedagogy; about the need for vigilance in educating children, crucial for the family and for the future of the state.Footnote 31 Furthermore, Simon Schama in Embarrassment of Riches argued that, whilst these images might well have been loaded with messages about the follies of the world, it is still significant that the artists drew on the subject matter of childhood.Footnote 32 His own analysis was predominantly concerned with paintings of children as expressive of the preoccupations of the young Dutch republic, caught between caution and ambition. To be Dutch, Schama proposed, was to be in a state of becoming, to be in ‘a sort of perpetual political adolescence’.Footnote 33 But this does not mean, he averred, that the images had nothing to do with the state of childhood itself; ‘that the world of Dutch children was swallowed up by the stereotypes of their parents. Suppose something like the exact opposite to be the case, namely, that adults shrank their own anxieties and pleasures to child size’.Footnote 34 Furthermore, Schama stressed, the didactic could rarely keep the ludic under control.Footnote 35
2. HISTORY, HISTORY OF ART AND VISUAL EVIDENCE
A thread running through many of these criticisms of Ariès’s use of iconography, particularly his assessment of medieval images, has been the claim that these artistic representations (or absences) tell us more about art than they do about childhood. It has been argued that Ariès, lacking a proper understanding of the various and complex ways in which images are created and convey meaning, drew an overly simplistic and direct link between pictures and society. He conceived of paintings as ‘peep holes’ through which to view the past; he understood them as proto-photographic snapshots. He applied, furthermore, a ‘scissors-and-paste’ approach to these images, and sheer accumulation took the place of analysis.Footnote 36 Ariès has thus been repeatedly attacked for falling into the trap so succinctly and chillingly rejected by Michael Baxandall: ‘One will not approach … paintings on the philistine level of the illustrated social history, on the look out for illustrations of “a Renaissance merchant riding to market” and so on’.Footnote 37 This takes us onto the wider, persistent methodological problem of the use of art as historical evidence; the fundamental question as to how, or indeed if we should use pictures to tell us about the past.
When Anita Schorsch published her Images of Childhood: An Illustrated Social History in 1979, she cheerfully went along with Ariès’s sense of a direct and unproblematic relationship between art and society: ‘[the imagery of childhood] has an immediacy that happily cuts across centuries to bring together the evidence of both historical documents and modern scholarship’.Footnote 38 However, criticism of Centuries of Childhood has often extended to a strident rejection of the value of imagery as evidence. This was the line taken by Linda Pollock in her important work, Forgotten Children, which proposed continuity in the place of Ariès’s change. Pollock rhetorically asked ‘how far do paintings represent reality?’, responding firmly; ‘there is no reason why there should be any connection between the representation and that which is represented’.Footnote 39 A number of critics have thus emphasised key artistic developments such as the advent of more broadly ‘realistic’ approaches, so that engagement with the distinctive qualities of childlike bodies becomes more to do with the embracing of a new naturalism in the Renaissance than anything specific to childhood. Artists in the fifteenth and sixteenth centuries did not begin to produce the kinds of image heralded by Ariès as indicative of a new awareness of the particular qualities of children because they began to look closely at and appreciate infants around them, but rather because they began to look closely at and appreciate antiquity. It was by turning to the Greeks and the Romans that they were able to create a vision of childhood with which we have sympathy. As Adrian Wilson put it; ‘the change [Ariès] … discovered – the growth of realism in the portrayal of the child – can be explained without reference to the mentalités in which he is interested. It is not the attitudes of people at large, but the forms and features of art, which changed.’Footnote 40 Art tells us about art, and that is all.
Erika Langmuir appears to have had this kind of attack as a backdrop to her discussion in Imagining Childhood. She strenuously distances her vast array of images of children, stretching from the Greeks through to the nineteenth century, from any historical claims whatsoever. She emphasises ‘the relative autonomy of art’, the crucial importance of cultural topoi and the dominant role of tradition.Footnote 41 Indeed, she explicitly sets herself in opposition to Ariès: ‘My aim is almost the reverse of his: not primarily to deduce attitudes to childhood from pictorial evidence, but to examine the imagery of childhood for what it tells us about images.’Footnote 42 Langmuir's project was largely conceived and undertaken at the Warburg Institute, and her scholarly analysis of certain tropes in art – presented in discussions which move from Greek art, to Ambrogio Lorenzetti in fourteenth-century Italy, to William Hogarth in eighteenth-century Britain, within the space of a page – is clearly indebted to the emphasis of that institution's founder on iconographic analysis through the juxtaposition of chronologically and geographically diverse imagery.Footnote 43
Most scholars writing in the wake of the first wave of criticism of Ariès’s treatment of art have not confined images of children so resolutely within the disciplinary boundaries of Art History. However, they have often firmly separated the History of Children and the History of Childhood. It is common to find these two areas set up as discrete entities at the outset of an article or monograph.Footnote 44 If interested in the History of Children, the author may well draw on demographic information and use sources such as diaries and letters. If concerned with the History of Childhood, s/he instead turns to poetry and novels; to paintings and engravings. There is a sense that as long as these two domains are kept apart – as long as one does not get confused between Wordsworth and the experience of a child in the early nineteenth century – then one's analysis is secure.
However, this persistent divide, I think, indicates a missed opportunity, sidestepping a challenge that is daunting, but surely worth engaging with. We should not stop trying to engage with the imagery of childhood as we continually develop and nuance our understanding of the history of children, just because Ariès’s use of pictures was admittedly rather crude. Certainly, we need to rid our analysis of any residue of a direct elision of the represented and the ‘real’. But there has been plenty of thoughtful engagement in recent years with how we might formulate the relationship between art and society, and how we might try to reconnect those categories in the wake of early, over-simplistic attempts to replace Formalism with the Social History of Art. Hugh Cunningham argued in favour of such reconnection in the introduction to his strategically titled Children and Childhood in Western Society since 1500, emphasising that ‘the need to be careful in the use made of pictorial evidence … does not render it useless, any more than it can be argued that the novels of Dickens tell us nothing about childhood in the nineteenth century’.Footnote 45 Cunningham admirably attempts to relate the two as one of the primary concerns of his book. He does, however, also assert that the story of Childhood is relatively easily pieced together, unlike the story of Children.Footnote 46 As this indicates, a persistent problem in the historiography has been the treatment of visual and literary representations of children as an overly homogenous entity, the particularities of different modes of expression insufficiently engaged with, and limited attention paid to competing idea(l)s.
But Cunningham's overall point is a crucial one, reinforced by Joanne Bailey in her recent study of parenting, which explores ‘how cultural values interacted with the experiential’.Footnote 47 When it is proposed that the reasons for changes in representations of children are to be found ‘in contemporary style, in artistic traditions, and in the functions of art’, there is often a sense that this is the end of the road.Footnote 48 Evocation of style and iconography are used to close down the possibility of valuable socio-historical significance. We cannot use Renaissance images of children because they are about the rediscovery of antiquity; we cannot turn to seventeenth-century Dutch paintings because they are loaded with symbolism and indelibly linked with the language of emblem books. However, proper engagement with style, traditions and artistic functions can restore the rich potential of these images as historical source material. This is not a case of using the specific disciplinary tools of the art historian in order to ‘translate’ the world of the visual: understanding the ‘distorting’ elements of painterly convention and iconographic tradition in order to remove them from the picture, and so convert the painting into something that might be more comfortably placed alongside ‘proper’ historical source material. We do certainly have to pay attention to a myriad of specific issues, such as the technologies of artistic production, the nature of patronage and the consumption of works of art – what Ludmilla Jordanova has described as ‘the processes through which representations come into existence’.Footnote 49 But those very processes can themselves be the subject of historical attention. Mediation should not be seen as something to work through: a barrier between the historian and the nugget of evidence on the far side of the picture. There is much within the process of mediation itself that proffers a rich source for historical enquiry.Footnote 50
3. PORTRAITS OF CHILDREN
I want now to focus in on one of the artistic genres featured by Ariès in his account of ‘the discovery of childhood’ in order to unpick some of these issues a little further: portraiture. Ariès proposes that it is unsurprising that we do not, at first, find portraits of children in the history of Western art. It is difficult to see why any parent would commission an image of an offspring who could well die before achieving maturity. When we start to see such pictures emerging in the fifteenth century, they give extra significance to an apparently novel interest in that stage of life, as high infant mortality rates persisted; ‘the new taste for the portrait indicated that children were emerging from the anonymity in which their slender chance of survival had maintained them’.Footnote 51 But Ariès is particularly concerned with the historical moment in which it became more common for children to be portrayed on their own: the late sixteenth and seventeenth centuries. This, he argues, showed a new perception of their importance, and a new awareness of and engagement with their distinct condition.Footnote 52 It is paintings such as Philippe de Champaigne's The Habert de Montmor Children of 1649, which reveal the arrival of ‘the modern family’ (Figure 2). Ariès uses this image in his discussion of dress, to illustrate the development of newly specialised costume for infants, evident in the long robes of the younger siblings and the two broad ribbons we see hanging from one child's shoulders, which came sartorially to designate the condition of childhood.Footnote 53 But new ideals are also evident in the artistic attention to offspring seen apart from their parents; in the monumentality of these children, brought close to the picture plane, engaging with us so directly; in their plump cheeks and pudgy hands; in the signs of sibling affection. It is in these elements that we most clearly recognise our own idea of childhood.Footnote 54
Philippe de Champaigne, The Habert de Montmor Children, 1649, oil on canvas, 122.5 × 186.4 cm. Museum of Fine Arts of the city of Reims. © photo: C. Devleeschauwer.
Ariès’s emphasis was well and truly on the seventeenth century. His story does give a significant role to the eighteenth century, as it is then that the family retreats from wider society and becomes more inward looking and private. But, he argues: ‘It was in the seventeenth century that portraits of children on their own became numerous and commonplace. It was in the seventeenth century, too, that the family portrait … tended to plan itself around the child.’Footnote 55 However, a key historiographical development since Centuries of Childhood was first published has been the relocation of the arrival of the ‘modern’ child in art to the eighteenth century – specifically to the mid and later eighteenth century – and, still more specifically, to Britain. Over the last few decades, the visualisation of this modern child has become emphatically associated with artists such as Henry Raeburn, Thomas Lawrence and, perhaps above all, Thomas Gainsborough and Joshua Reynolds.Footnote 56 This was the narrative established by J. H. Plumb in his 1975 article on ‘The New World of Children in Eighteenth-Century England’:
Up to about 1730 family portraits are formally posed groups; increasingly, however, after 1730 children are shown playing or reading or sketching or fishing or picknicking with their parents … Also, portraits of individual children are far more common in the eighteenth century than in the seventeenth, again arguing both for a change in fashionable attitudes, and also, maybe, for a greater emotional investment in children by parents.Footnote 57
Desmond Shawe-Taylor deployed the same argument and chronology in The Georgians in 1990, but the most significant publication accompanied a 1995 exhibition held at Berkeley, entitled The New Child: British Art and the Origins of Modern Childhood 1730–1830.Footnote 58 In this, James Steward begins by noting the long tradition of representations of children in European art, but also that this ‘generally lacked both a sophisticated sense of the individuality of the child and any real interest in the child's psychology or even experience as distinct from that of the adult’.Footnote 59 There are glimmers of the incipient New Child in the early 1700s, but it is only once we get to the mid and later eighteenth century that we find paintings that truly engage with the spontaneous and free world of childhood. This is Steward's discussion of Reynolds's portrait of the two-year-old Francis Hare, executed in 1788 (Figure 3):
The child now dominates the canvas and the natural setting. No longer vulnerable, the child is a symbol of confidence for the future … It is to the artist's credit that here we forget that this is a child seen through adult eyes, for he is presented in his world without filters, without emblems … Both Reynolds and Gainsborough underline the validity of this child's world, giving it descriptive elements appropriate to childhood.Footnote 60
Joshua Reynolds, Master Hare, 1788, oil on canvas, 77 × 64 cm. © photo RMN-Grand Palais (musée du Louvre) / Michel Urtado.
The relocation of ‘the new child’ to mid and late eighteenth-century British art has taken place with relatively little comment on the change in temporal boundaries and the delimitation of Ariès’s geographical framework. Emma Barker, in a recent article on Jean-Baptiste Greuze's Girl with a Dog (1767), notes with some surprise that historians of eighteenth-century French art have not engaged with Ariès’s narrative to anything like the same degree as those concerned with Britain.Footnote 61 It is also this new version of Ariès’s story that has underpinned and been disseminated in accounts aimed at a wider audience. Reynolds’s cherubic infants, together with Gainsborough's evocative portraits of his daughters, are popular choices on object lists for exhibitions such as Pictures of Innocence, held at the Holburne Museum in Bath in 2005.Footnote 62 Anne Higonnet drew on Steward when she published her book of the same name with Thames and Hudson in 1998. In her account, the ‘Romantic child’ of late eighteenth-century British portraiture introduces the paradigm of childhood innocence, the history and development of which she then tracks through the nineteenth and twentieth centuries.Footnote 63
But Ariès never mentions Reynolds or Gainsborough. For him, the crucial names are those of Anthony van Dyck, Peter Paul Rubens and Philippe de Champaigne.Footnote 64 Those names are, however, now usually pushed back into the ‘before’ category in the oft-rehearsed ‘before’ and ‘after’ narrative of the pictorial emergence of ‘the new child’. Thus, whilst Ariès presents van Dyck as one of the illustrators of modern childhood, Steward describes The Five Eldest Children of Charles I as showing almost negligible expression and characterisation, the ‘essentially rigid poses of the bodies, the detailed costumes, and the aloofness of the figures’ making it clear that evidence of wealth, status, regality and dynastic succession are the sole artistic concerns (Figure 4).Footnote 65 Van Dyck is given the same position on the evolutionary spectrum in Higonnet's book, when she similarly emphasises the adult poses and costumes featured in portraits such as The Villiers Boys (1635).Footnote 66
Anthony van Dyck, The Five Eldest Children of Charles I, 1637, oil on canvas, 163.2 × 198.8 cm. Royal Collection Trust. © Her Majesty Queen Elizabeth II 2016.
Van Dyck has thus often been a casualty of the shift in interest to the later period. However, more sensitive analyses of his portraits of elite and royal children have rather emphasised a subtle play. Certainly, his portraits had specific functions. They were used to emphasise status and the security of lines; they helped to establish and maintain alliances, playing a crucial role in marriage negotiations; the main concern was arguably with what the child would become, rather than what s/he was at that moment.Footnote 67 The consequent visual language, emphasising elaborate costumes, meaningful attributes and dignified poses, might not seem to tell us much about the particular world of the young sitters depicted. However, these pictures do tell us something about childhood. Certainly, dynastic portraiture focuses on a particular aspect of the sitter's identity, which it was necessary to select, enhance and perhaps exaggerate in a genre routinely associated with the desire to ‘put one's best foot forward’ – but it does, at least, tell us about that aspect.Footnote 68 Dynastic concerns were very much part of the lives of elite children. Moreover, van Dyck, like other portraitists of this period such as Velasquez, appears to have been interested in the relationship between the functions of dynastic portraiture and the particular physicality and experience of the child.Footnote 69 Young Charles may wear the apparel of a grown man, but his face, with its rounded cheeks, is very much that of an infant. His leaning posture may echo that featured in innumerable adult male portraits by the same artist, but the position of his elbow is slightly higher than is the norm, even uncomfortably so as he rests his arm on the head of the huge mastiff, whose scale reinforces our sense of the boy's diminutive form. Van Dyck creates an identifiable space between the bodily capacities of the child and the formal, dynastic purposes of the image in which he is depicted. He also invites us to contemplate the process by which Anne on the right – a chubby, barely clothed baby, only prevented from falling by her sister – will become a more mature girl, like Mary on the left. If the emphasis of dynastic portraiture is on what the infant is to become, then van Dyck's painting does not so much project the child into the future as explore the active process by which the child is transformed into the adult. It is, I would argue, such complex play between the state of a child and future adulthood that allows van Dyck's work to occupy its notably mobile position in the story of ‘the discovery of childhood’.
4. MASTER CREWE
In the final section of this article, I want to turn to one of those eighteenth-century British portraits often held up as revealing the new awareness of and delight in the idiosyncrasies of children prioritised by Ariès: that of Master John Crewe, painted by Joshua Reynolds (Figure 5).Footnote 70 Reynolds is, as I have noted, one of the most prominent artists in accounts of the rise of ‘the new child’ in portraiture, and Master Crewe as Henry VIII is one of the most famed and reproduced examples of his work.Footnote 71 I want to use this painting as a case study: to explore what such an image might be able to tell us about ‘the child’ in this period; to help scrutinise the idea that these portraits indicate a novel sense of the import of infants in both the family and society at large, together with a ‘modern’ understanding of the distinctiveness of their condition; and, finally, to help unpick that fraught issue of the complex ties between popular ideals of childhood and particular lives.
Joshua Reynolds, Master Crewe as Henry VIII, c. 1775, oil on canvas, 139 × 111 cm. Private collection, on long-term loan to Tate Britain. ©Tate, London 2016.
Master Crewe was executed in c. 1775, conceived as a pair with a portrait of John's older sister, Frances (Figure 6). Whilst the pendant remained unfinished due to the girl's sadly premature death, John, three years old, is a stridently robust figure, as fine a candidate for the title of ‘new child’ as any. His figure has the plumpness characteristic of children of that age, his cheeks are round and ruddy, his hands are fleshy and dimpled. He is also splendidly pleased with himself, as he assumes the costume and stance of Henry VIII as immortalised by Holbein.Footnote 72 So that we are fully aware that this is a child's game of dressing up, Reynolds has included the boy's own shoes and his green coat with its large round buttons, recently removed and draped over a stool. This is a comfortable garment, designed specifically to accommodate a child's body and activities, and its presence enhances the joke of the portrait – drawing attention to the gap between what a little boy in the mid-1770s would typically have worn, and the historic adult costume he has donned.Footnote 73 Furthermore, one of the spaniels, a late addition to Reynolds’s composition, sniffs his master's leg in apparent suspicion at his altered appearance.
Joshua Reynolds, Miss Crewe, c. 1775, oil on canvas, 137 × 112 cm. Private collection, on long-term loan to Tate Britain. ©Tate, London 2016.
This is an image of a small boy delighted with his game of dressing up – but there is much at stake here concerned with artistic issues and conventions that had little to do with John at all. Master Crewe is a learned and witty painting that reveals Reynolds’s ambition as an artist, his drive to elevate the genre of portraiture (often disparaged as merely concerned with reproducing likeness, tempered only by the drive to flattery), and his ability to engage with complex allusion and quotation.Footnote 74 This picture is a likeness of the child who visited the artist's studio in Leicester Fields, but it is also a painting in a mock-heroic tradition, admired by Horace Walpole in significant terms: ‘Is not there humour and satire in Sir Joshua's reducing Holbein's swaggering and colossal naughtiness of Henry VIII to the boyish jollity of Master Crewe?’Footnote 75 Walpole's statement emphasises that Holbein's portrait of the king has been turned into the image of a young boy: Henry's well-known swaggering demeanour has been punctured by comparison with a toddler. Holbein's painting, rather than the sitter, is the starting point in this account. Indeed, Reynolds’s portrait not only pays tribute to a famous sixteenth-century image, it is also a clever, humorous reflection on his own much noted habit of such ‘borrowing’ from other artists. Master Crewe was placed on display at the Royal Academy the year after the artist had been attacked by his fellow painter, Nathaniel Hone, as a ‘Conjuror’, who would delve freely into Old Master prints to lift poses and compositions for his pictures.Footnote 76 In Reynolds’s view, this was legitimate, undertaken in a spirit of emulation, even competition, and was not a matter of plagiarism – but his decision to give his young sitter, in effect, the job of copying Holbein, rather than making it purely his business as the artist, is a clever response to such criticism.Footnote 77 It is also significant that the Old Masters that Reynolds preferred were those of the Roman High Renaissance, Raphael and Michelangelo, artists famed for their ‘grand manner’. Here, for the immature form of a child, he has instead pointedly referenced an artist linked with a style he routinely put firmly in its place as detailed, mimetic transcription of the physical world. He has thus aligned the undeveloped state of childhood with a mode of painting he repeatedly identified as limited in scope and ambition.
This analysis helps to show up, I think, the problem with the idea that the basic fact of an increasing number of child sitters being represented independently, on their own canvases, revealed a new interest in their individuality.Footnote 78 To a considerable degree, the child isolated in this way, even if accompanied by siblings, was actually able to function as more of an artistic ‘vehicle’. It was easier for those long traditions by which children have been united with symbols, with metaphor and allegory, to be activated and maintained, if they were separated from adults, and removed from the context of the wider family group. Thus, whilst a key element of Steward's narrative was that the later eighteenth-century ‘new child’ was freed from emblematic or iconic baggage, able to be more ‘him/herself’, that baggage clearly remained, and what Marcia Pointon has noted as the child's ‘particularly powerful connotative capacity’ was undimmed.Footnote 79
The way in which Reynolds represented Master Crewe thus has much to tell us about the painter, and artistic issues in this period more broadly. However, I would refute that this might in any way neutralise the image's potential to function as an historical document, with potential value for the history of the child. A portrait does not only concern the portraitist: it also concerns the sitter, the patron, its anticipated audiences, and it has, therefore, to fulfil a variety of objectives. Master Crewe was an overweening display of Reynolds’s abilities as an ingenious and erudite artist, a treatise on the practice of artistic referencing, but it does also show a clear, and relatively novel, engagement with the physical and behavioural characteristics of childhood. If ‘putting one's best foot forward’ in earlier dynastic portraiture had meant a dominant emphasis on wealth and future high status, then, in this era, it meant engaging with the particularities of the condition of infancy, presented in such a way as to make it clear that the viewer of the picture should find those particularities appealing. Reynolds, after all, had to operate in a market, securing and then maintaining a popularity that could underpin his increasingly high prices.Footnote 80
Here, we do need to engage with developments in broad ideas of childhood; above all, with the ways in which those ideas were affected by the culture of sensibility in this period. Increasingly dominant in Britain from the mid-century onwards, sensibility gave emotions a new status and significance, so that commentators came to laud their direct expression, whilst elevating values associated with Nature.Footnote 81 The impact of sensibility on ideas about childrearing came through most clearly in the writings of Jean-Jacques Rousseau, and especially his famous treatise Emile, translated from French into English in 1763.Footnote 82 Although Rousseau's role can sometimes be overemphasised – the sense of a dramatic transition in attitudes occurring at the precise moment this text appeared doing disservice to the work of other writers who paved the way – his philosophy did greatly help to make the image of the ‘natural’ child, evident in Reynolds's painting, very fashionable.Footnote 83 In the portrait of Master Crewe, we see a comfortable coat, allowing growth and easy movement, and indulgence of childish play, as well as the robustness and confidence which Rousseau sought to encourage (and which was notably gendered, highlighted here by the contrast with the bashful posture of John's sister). Lacking additional concrete evidence, it would be hard to assert without a shadow of a doubt, that John's parents, John and Frances Crewe, fully subscribed to Rousseau's ideals in their attitude to their son, and in their approach to his upbringing and education. However, it is safe to say that they wished to be seen as subscribing to those ideals, by commissioning a portrait in which indulgence of childish physique and play are so readily evident.Footnote 84 We rarely have definitive evidence of the relative inputs of patron and artist into the conceit and appearance of a portrait, but the simple fact that the Crewes turned to their good friend, Joshua Reynolds, for this commission tells us much. To choose Reynolds to execute a portrait of one's child(ren) was to choose the kind of imagery in which he specialised; a way of depicting infancy that could be praised by a reviewer of one of his portraits, exhibited in 1784, as ‘the genuine transcript of nature’.Footnote 85
Images of children are distinctive within the broader genre of portraiture as the person ‘putting their best foot forward’ is not really the sitter.Footnote 86 It is the parent, or parents, who usually commission the work, and who are primarily associated with the ideals embedded in the picture. And Master Crewe not only projected John and Frances as in accordance with powerfully fashionable ideals of childrearing to those who had access to their home. This painting also spoke to those visitors who saw it in the Royal Academy exhibition of 1776, and knew that the ‘Portrait of a boy in the character of Henry the Eighth’ listed in the catalogue was a likeness of Master John Crewe.Footnote 87 Those visitors could also have been amongst those who purchased the mezzotint after the portrait, executed by John Raphael Smith and published by John Boydell, already available in the print shops when the painting was placed on display at the Academy.Footnote 88
However, considering the wider audience of the portrait takes us back to Reynolds, who not only had to please his wealthy patrons, but also to consider his broad appeal and reputation amongst the thousands who attended the exhibitions, those who formed the national and international markets for the prints so widely distributed after his pictures, and the critics who opined on his output. Many of those who saw Master Crewe on the wall at the Academy would actually have had little sense of or even interest in the sitter's identity. The same is true of those who inspected and bought a copy of the Smith mezzotint. Crucially, Reynolds’s child portraits could be deemed pleasing and affecting as generic images of infancy, regularly praised as ‘sweet’, ‘natural’, and playful’.Footnote 89 His image of Frances Gordon ‘in different views’, for example, was lauded in 1787 for the ‘sentiment, expression, and clearness, and warmth of colouring, that all must feel, but which perhaps the President of the Royal Academy alone can describe’.Footnote 90 The key phrase here is ‘that all must feel’: this is about the way in which childhood, in an era that prizes the emotions, can affect the adult spectator; the pleasurable sensations it is capable of creating in a viewer.
And this, arguably, is at the heart of issue. Sensibility did not just affect values around childhood and parenting; it also greatly developed the potential of an image of a child as a likely prompt to the experience and display of feeling, and this had long-lasting impact on cultural trends. One of the key points made in criticism of Centuries of Childhood is that Baldick's translation of Ariès’s ‘sentiment’ as ‘idea’ loses much of the original meaning.Footnote 91 ‘Sentiment’ in French denotes a sense of feeling, as well as a concept, and much of Ariès’s argument about childhood is about his (and what he presumes also to be his reader's) emotional engagement with the condition. When Ariès looks at medieval pictures, he observes that he does not find the ‘grace’ or the ‘rounded charms’ of later works.Footnote 92 When he charts developments in artistic representations of the Infant Christ, he states that they became both more ‘realistic’ and more ‘sentimental’ in the twelfth and thirteenth centuries. As we move through their inexorable evolution, we find evidence of still more engagement with the ‘graceful, affectionate, naïve aspects of early childhood’, and we begin to detect ‘sentimental realism’.Footnote 93
It has been rightly pointed out that this is fraught territory. One aspect of the familiar accusation of ‘present-centredness’ that has often been levelled at Ariès is the danger of projection when dealing with such an emotive topic as childhood.Footnote 94 To quote Jordanova again: ‘Our capacity to sentimentalize, identify with, project on to, and reify childhood is almost infinite.’Footnote 95 Pointon has also noted the temptation and tendency for scholars to shift into the present tense when describing the charms of Reynolds’s children, and I did not avoid the trap when writing about Master Crewe, above.Footnote 96 Such projection, such visceral response, can mislead us into thinking that our engagement with childhood constitutes realisation of a ‘truth’, and the consequent deduction that this was a ‘truth’ first appreciated in the eighteenth century. It is certainly a temptation that has to be avoided – but the existence of that temptation provides its own historical insight. Perhaps one of the most novel ideas about childhood that developed in the eighteenth century was that it could, and indeed should, be a prompt to an emotional reaction. Ariès was preoccupied with the ‘grace’ of images that fully engage with the distinctive qualities of children, and this is also an emphasis which Barker has noted as characteristic of eighteenth-century commentaries by writers such as Claude-Henri Watelet. In Watelet's view, childhood and youth are the ‘ages of grace’ because the body honestly expresses the movements of the soul in a free, easy and simple way.Footnote 97 Perhaps above all it is the new emotional and aesthetic response to the child that emerged in this period, which has led us to identify it as witnessing the emergence of ‘modern’ childhood.
The expressive potency of images such as Master Crewe also points to the peculiarly direct power of the visual. It is vital to look at Reynolds’s portraits alongside the literary edicts of writers such as Rousseau – within the context of ideals expressed in a variety of media – but it is the artist's depiction of infantile simpers, wayward curls and large round eyes that triggers the kind of comment about which Pointon complains, and which it can be hard to resist. This highlights the need to tease apart various cultural constructions of childhood, and also the imperative to acknowledge the degree to which they formed as well as reproduced meaning. Reynolds’s visualisation of the physiognomy, the bodies, and the behavioural characteristics of children in the public arena of later eighteenth-century art had a particular force, and contributed to widespread ideas about childhood. In the period, it was often noted that sight was the most direct conduit to the emotions, and that the artist had the advantage of an immediate effect that was denied to the writer, who had to unfold a description.Footnote 98 Reynolds’s portraits, furthermore, amplified those aspects of the ideal of childhood that indelibly or potentially had a visual component, and helped to create an increasingly recognisable pictorial language of appealing infancy. Paintings such as Master Crewe should thus be seen as adding to and inflecting literary ideals, rather than as merely creating an image of an indulged child to be set alongside texts such as Emile.Footnote 99
5. CONCLUSION
This discussion of cultural ideals, Rousseau, sensibility, and the child as a prompt to a pleasurable emotional response, has taken us a long way from the particular experiences of little John, aged three. But the point to which I would like to return, and finish on, is that the place of the child in eighteenth-century sensibility should not mean that we restrict such portraits to the History of Childhood, even if that is nuanced with due attention to the particularities of Reynolds’s work and the distinctive place of the visual. These pictures do present business for the iconographer and require engagement with the history of Western art. Master Crewe was clearly about the artist's desire to lay claim to an ambitious, engaging art, which erudite viewers could appreciate. It was also a prompt to an emotionally charged response, which needs to be understood through broad engagement with cultural values of childhood in circulation at the time. But this is still a portrait, executed for John's parents, who paid a considerable amount of money for it.Footnote 100 It remained in the family's possession, and is only now on display at Tate Britain thanks to a long-term loan arrangement. It is a portrait for which John Crewe would have had a number of sittings, visiting Reynolds at his studio, and that experience is embedded in the conceit of the picture.Footnote 101
In his quotation of Holbein, Reynolds was certainly playing with the idea of artistic referencing, but he was also engaging with that tension opened up by Anthony van Dyck between the child's experience of sitting and the construction of the portrait. The act of self-presentation is here aligned with childish masquerade and love of posturing, so that the necessarily contrived pose taken from the famed image of Henry VIII – gloves grasped firmly in the right hand; left hand hooked into the cord from which a dagger hangs – becomes something with which the boy actively engages as a game, rather than something by which he is constrained. Anecdotal evidence from the likes of James Northcote, pupil to Reynolds for five years, intimates that the success of his portraits of children was founded in an ability to engage with the distinctive, potentially difficult dynamics of a child's sittings.Footnote 102 Some of this is certainly apocryphal, but these stories are sufficiently plentiful to suggest an artist who was interested in being inspired by, rather than wrestling to manage, the business of painting children. William Russell, cowering in the corner of Reynolds’s painting room out of ‘a horror of being painted’, was apparently transported into a portrait of his siblings engaged in taking on St George's dragon.Footnote 103 Lady Anne Spencer shrank from Reynolds’s scrutiny in the studio, but he used the pose to depict her withdrawing from a grotesque mask held up by her sister in the famous Marlborough Family.Footnote 104 Yet the portrait of Master Crewe is distinctive as no ‘conversion’ or translation has taken place. What we see here is the sitting; the child having cast his coat aside, donned his fancy dress, presenting himself to the painter.
Miss Crewe's proud but decidedly self-conscious expression, lips tightly pressed together, also speaks of the experience of being a child faced with the gaze of the most famed portraitist of the day. But Frances's portrait is most indelibly bound up with her premature death, as the poignant fact of her demise is embedded in its detectably unfinished surface. Frances's sketchily blocked in gloves, together with the emphasis on her black silk calash, raised against the cold, and the wintery landscape with its frozen sky, take on melancholic significance as a result of this biographical insight, and the themes of transience and fragility which run through so many images of childhood become concrete and particular. There is even a suggestion that Reynolds did not ask her parents, as longstanding patrons and friends, for payment out of respect for their loss.Footnote 105 This portrait is intimately caught up with the premature death of this little girl. The melancholic aspects of the cultural image of childhood in this period must always take us back to the individual losses and anxieties that underpinned it.Footnote 106
The ‘real’ child is there, in these portraits. S/he is there in the physical and social encounter of artist and sitter; in the commissioning of the painting by the family, and its subsequent display; in what her/his parents wanted to see in and have advertised by the picture. If we are concerned with actual historical actors, then the fact the Crewes engaged with contemporary ideals of childhood in the expensive portrait they commissioned of John, which they allowed to be exhibited and engraved, before hanging it in their home, tells us a great deal. And, although we still need to develop our methodologies and models for the exercise, it is well worth considering the relationship between the patrons’ embracing of distinctively eighteenth-century ideals of childhood in this portrait and its pendant, and their perception and actual treatment of their son and daughter – the interaction between cultural trends on the one hand, and expectations, standards and practices on the other. The relationship is certainly neither direct nor straightforward, but it clearly exists, as scholars such as Bailey have emphasised.Footnote 107 Portraiture is, I would argue, a particularly profitable site in which to bring together the History of Childhood and the History of Children.
ACKNOWLEDGEMENTS
I first presented this material at a seminar organised by the Centre for the History of Childhood at the University of Oxford, and am grateful to the organisers, Laurence Brockliss and George Rousseau, as well as the participants, for their valuable feedback. I would also like to thank Emma Barker and the three referees for Continuity and Change for their insightful comments on an earlier draft of this article.
