Visual ideology, as developed in the work of Nicos Hadjinicolaou (1978), is an approach to images that prioritises iconology in its ideological deconstruction. Hadjinicolaou developed his approach as a counter to what he saw as the reactionary approach of art history. His thesis and the methodology he developed, although about fine art painting can be applied to any constructed image.
Hadjinicolaou reasserts the Marxist thesis that all history is the history of class struggle; and that also applies to painting, which most art history ignores. He contends that art reflects the ideologically distorted perception of people's real existence: ideology functions to hide the true nature of class relations through the construction of a plausible perspective on life that perpetuates an idea of the unity of mankind and a positive worldview.
Hadjinicolaou (1978, pp. 95–6) defines visual ideology as:
A specific combination of the formal and thematic elements of a picture through which people express the way they relate their lives to the conditions of their existence, a combination which constitutes a particular form of the overall ideology of a social class.
Visual ideology is not apparent in painting; it does not jump out and demand attention, any more than ideology in every other part of lived experience is obvious. It needs to be uncovered by digging beneath appearance. This is a difficult and time-consuming process. Essential to this process is analysing art by situating it in its historical milieu.
In this, Hadjinicolaou reflects the concern with context, although this is not just the immediate in which the image is produced but the social historical context that informs the actions of the image maker. A literary example would be why William Shakespeare's Julius Ceasar does not conclude with Ceasar's assassination but sets it half way through the play and focuses on the comeuppance of the regicides in the remainder. One might argue a pragmatic move by Shakespeare given the power and ruthlessness of the Elizabethan monarchy and the various contemporary threats and attempts on the life of the sovereign.
Hadjinicolaou argued that each class (or layer of class) ought to have its own visual ideology but in the main the ruling class visual ideology has been so dominant that they have permeated the visual ideologies of the non-ruling classes, rendering as undeveloped the visual ideology of such classes. He argued that the history of the production of painting has been the history of ruling class ideologies (a point also made by Roger Taylor (1978)). Nonetheless, Hadjinicolaou argued for the change potential of visual ideologies, rather than the disavowal of art proposed by Taylor.
Hadjinicolaou, argued that visual ideology is not restricted to the content of a painting but also in the form. Thus he denied the art historical idea that one can solely analyse a work aesthetically by focusing on the formal elements; in short he rejected the traditional art history separation of form and content. He suggested that semiological concepts such as sign, signified and signifier are useful in combating the distinction between form and content and thereby aiding in the analytic work aimed to 'relate visual ideology to other types of ideology' (Hadjinicolaou, 1978, p. 144)
Visual ideology inheres in the style of a painting; evident in both form and content. Hadjinicolaou, (1978, p. 94) thus prefers the more analytic term visual ideology rather than style. For him, every painting should be analysed to see how 'the formal and thematic elements of a picture are combined on each specific occasion'. Such a combination is a particular incidence 'of the overall ideology of a social class' (reached through knowledge of the literary, aesthetic, religious, political economic and other ideologies of the class). There is a dialectical relationship between class ideology and visual ideology as the former informs the latter but the latter is a specific instance that illuminates the historically specific class ideology. Neither can be directly deduced from the other.
Apart from the use of style to designate a non-individual (collective) phenomenon, only the conception of 'picture's style' can be retained. In this sense one may speak of the visual ideology of Rubens' Marie de Medici Landing at Marseilles (Louvre) and its relation with 'baroque style'. Thus both the style of a particular work and the style prevailing in a region or on a nation-wide scale can only be approached through the study of their relationship with contemporary collective styles. On the other hand, a collective visual ideology (as for example the 'baroque style') can only be studied seriously by means of an analysis of individual pictures. Hence the art historian must again pursue his research within an apparently viscous circle, as he [sic. Hadjinicolaou, in the preface to the English edition subsequently apologised for not realising he had exclusively used the male pronoun] did when investigating the relationship between an overall class ideology and a visual ideology. Research into the style of an individual picture cannot be carried out in isolation from the collective visual ideology to which it belongs, but the collective visual ideology cannot be ascertained without reference to individual pictures. (Hadjinicolaou,  1982, p. 246).
That a painting reflects a visual ideology does not mean that there is a visual ideology of the artist. An artist may be part of a group in which there is a visual ideology or be located within a political system that evinces a visual ideology but the artist can move and so not all paintings by an artist exhibit the same visual ideology.
Hadjinicolaou argued that, similar to ideology, there are both positive and critical (negative) visual ideologies. Positive visual ideology reaffirms dominant social class ideology: for example, glorifying religious and political ideologies, via such things as allegorical painting. An example is Peter Paul Rubens' Rape of Ganymede (c. 1636). Critical visual ideology engages with dominant ideology through its treatment of the subject, including the form as well as the treatment of the content. An example is Rembrandt's The Abductionof Ganymede (c. 1635).
The two paintings were representations of Ovid's story (in Metamorphoses, 10 pp. 155–61) of Jupiter, transforming himself into an eagle, to carry off the shepherd boy Ganymede to Mount Olympus, where he became Jupiter's cup-bearer. Both paintings depict the moment of abduction.
Rubens' painting (Figure 1) has a diagonal (bottom left to top right) ascending composition. There is a lightning bolt visible among the clouds, symbolising both the god and the force of the abduction. The youthful figure of Ganymede is classical, based on one of the children of the Hellenistic group of Laocoonte in the Vatican Museum. The youth looks apprehensive (perhaps rapturous) and yielding.
Figure 1: Rubens The Rape of Ganymede
Rembrandt (Figure 2), however, depicts the beautiful shepherd, as anything but attractive. The figure of Ganymede is that of a plump, crying, terrified, reluctant child, whose limited amount of clothes are almost being pulled off, as it is being carried off by the eagle. The image clearly suggests the abduction of a young child: the background and the sky are dark and foreboding. The painting has a diagonal (top left to bottom right) descending composition. It seems Rembrandt's painting is a disparagement of the classical mythology (and of those who invoke its voyeurism as respectable).
Figure 2: Rembrandt The Abduction of Ganymede
Although most art works exhibit positive visual ideologies an individual picture, which may belong to a collective visual ideology (by dint of subject matter or style), may simultaneously reflect a critical visual ideology. This is evident in Rembrandt's The Abductionof Ganymede: a classical subject but a critical treatment that has disturbed bourgeois art critics. For Hadjinicolaou, having both a dominant (positive) and critical (negative) visual ideology on the same canvas enables the viewer to perceive the dominant ideology and to understand the contemporary class struggles of the period they were painted.
Arguably, some paintings have a critical visual ideology that overwhelms the dominant visual ideology that may have initially framed the content. William Holgarth's satirical paintings are a case in point. His depiction of an upper-class arranged marriage, Marriage A La Mode, in six paintings starting with the contract (Figure 3) and ending in the death of the couple is clearly an attack on wealth and privilege. At first glance the first painting in the series looks conventional but on inspection the critique becomes abundantly clear.
Figure 3: Hogarth, Marriage A La Mode(The Contract)
The marriage between the son of the bankrupt Earl Squanderfield with the daughter of a rich merchant is being arranged. The earl is sitting down with his legs bandaged, which is a sign of gout. This disease, which is caused by eating rich and fine foods only the wealthy could afford, is used by Hogarth to emphasize the status of the earl. The earl himself is keen to showcase his position in society, by pointing out his family tree. The merchant, plainly dressed, is handing over the marriage contract. To them, this is just a business arrangement. The merchant is selling his daughter to further his position in society, while the earl needs the money to build his new house and replenish his finances.
Behind them, sitting on the sofa, are the future bride and groom. The son of the earl, richly dressed, isn't remotely interested in the affair. He doesn't even bother to look at his fiancée, busy as he is at staring at himself in the mirror. The poor young girl, on the other hand, is distraught. She's wringing her handkerchief, while the lawyer Silvertongue is trying to console her. Over the couple hangs a picture of the Medusa by Caravaggio, while, in the left bottom corner, two dogs are chained to each other. These elements further emphasize how doomed their marriage is. (History And Other Thoughts, 2013).
Various critics have taken issue with the visual ideology approach. John Berger, (2000, p. 202), for example wrote:
There is nothing in common, he [Hadjinicolaou] says, between a Louis David portrait painted in 1781, the David painting of 'The Death of Marat' of 1793, and his painting of 'Madame Récamier' of 1800. He has to say this because, if a painting consists of nothing but visual ideology, and these three painting clearly have different visual ideologies reflecting the history of the Revolution, they cannot have anything in common. David's experience as a painter is irrelevant, and our experience as spectators of David's experience is also irrelevant. And there's the rub. The real experience of looking at paintings has been eliminated.
However, in discussing aesthetics, while denying the existence of any aesthetic effect independent of ideology or historical context, Elizabeth Chaplin (1994, p. 65) noted that Hadjinicolaou claimed that:
there is a correlation between the pleasure felt by the observer when he recognises himself in a picture's visual ideology. And since that picture may contain different values it may appeal to one person and not to another. Yet aesthetic judgements which evaluate works of art differently and which pronounce on their 'aesthetic' effects or 'beauty' always derive from the aesthetic ideologies of social groups. Aesthetics, per se, is doomed because it is a discipline without subject matter: '"What is beauty?" or "Why is this work beautiful?" must be replaced by the materialist question "By whom, when and for what reasons was this work thought beautiful?"' (Hadjinicolaou 1978:183).
For Hadjinicolaou, visual ideology does not just involve reconstitution of the historical setting of painting, the wider ideological context and the way the artist responded to it. On the contrary, visual ideology involves an act of interpretation. Hadjinicolaou's approach is informed by hermeneutics, which sees interpretation as an active pursuit that adds something to the reality to which the interpreted object belongs. Analysis of the work of art requires situating it within the wider social, historical and ideological context. This means taking account of ideology of both the genesis and interpretive stages. In so doing, interpretation creates an irrecoverable distance between moment (context) of genesis and the moment of interpretation. Thus, a work cannot be reduced to a single meaning. Rather, Hadjinicolaou considers that a work has no meaning except that which is the outcome of a tension between several incompatible meanings.