Cloud [draft] [#digitalkeywords]

“The web is full of beatific images of laptops sitting on heavenly clouds.  The rhetoric of the data cloud likes to exploit the peaceful, inconsequential parts of the tradition while suppressing the rest. ‘The cloud’ is a huge PR achievement for the IT industry, but it is profoundly deceptive.”

The following is a draft of an essay, eventually for publication as part of the Digital Keywords project (Ben Peters, ed). This and other drafts will be circulated on Culture Digitally, and we invite anyone to provide comment, criticism, or suggestion in the comment space below. We ask that you please do honor that it is being offered in draft form — both in your comments, which we hope will be constructive in tone, and in any use of the document: you may share the link to this essay as widely as you like, but please do not quote from this draft without the author’s permission. (TLG)


Cloud — John Durham Peters, University of Iowa

Nothing is quite so packed with meaning as clouds.  Despite their reputation as flighty and insubstantial, clouds have carried a wide range of discourses, practices, and arts for a very long time.  Or perhaps it is precisely their apparent blankness, mutability, and vanishing mode of being that makes them such a ripe canvas for human creativity and criticism.  The term “cloud” has earned its place as a digital keyword because of its widespread current use for server-based online data storage, but the nearly instant and universal acceptance of this term would be impossible without the much longer legacy which I trace here.

“Cloud” is etymologically related to “clod” and the Oxford English Dictionary reports that the first but now obsolete meaning of “cloud” was rock or hill.  Perhaps the atmospheric rather than geological sense of the term, emerging in thirteenth-century English, was originally a metaphorical projection of terrestrial to celestial cumulus rather like the more recent usage.  Tracing the comparative pathways of cognates shows how closely sky and clouds have been associated.  “Sky” in Norwegian, a cognate to the identically spelled English term, actually means “cloud,” and “welkin” in English, an archaic term favored by Shakespeare and other poets that means “sky” or “celestial vault,” is cognate with the Dutch “wolk” and the German “Wolke,” both of which actually mean “cloud.”  The deep association of clouds with the upper sphere or celestial realm has conditioned much of the history of their meaning.  They are also found in the night sky such as the “Magellanic clouds” (dwarf galaxies) visible in the Southern Hemisphere, or interstellar clouds of dust and gas known as nebulae, from the Latin clouds.  To be “in the clouds” has long meant to be in the sky, and by implication, to be in a fanciful, mystical or “ungrounded” state.  In the digital “cloud,” the sense of whim, instability or risk seems remarkably absent, perhaps enforced by the consistent use of the singular “in the cloud” to contrast with “in the clouds.”

Throughout its varying history “cloud” has always meant an agglomeration or amassing of materials, whether of stone, water vapor, or data.  Thus the koinē  Greek νέϕος μαρτύρων “cloud of witnesses” (Heb. 12:1) was solidified in English via the King James Bible (1611).  In Paradise Lost (1667) John Milton mentions a cloud of locusts and most intriguingly, the OED supplies the 1705 phrase “a cloud of informations.”  Clouds could be crowds or swarms of arrows, flies, or birds, anything bunched that can cast a shadow.  This sense also extends to use of cloud as a verb in English since the sixteenth century.  “To cloud” means to cover with darkness, obscure, or dim, and can have related sense of ill humor or gloom as in “cloud of suspicion” or “under a cloud.”  Such negative meanings, as we will see, seem to have little currency in the digital “cloud.”

Perhaps the oldest discourse around clouds is a theological one.  In Homer “νεϕεληγερέτα” is an epithet for Zeus (“he who collects the clouds”); clouds can be just as important attributes of divinity as thunder and lightning.  In the Hebrew Bible, YHWH has his habitation in the cloud, and guides the people of Israel on their desert sojourn with a pillar of cloud, a sign that both obscures his presence but also thereby points to it.  In the New Testament, Christ is said to return to the earth “in the clouds,” a theme beloved of Baroque painters, and in popular culture, there are innumerable images of a cloud heaven inhabited by God, saints, and angels.  Due to their association with the celestial realms, clouds are ready metonyms for deities of all kinds.

A counter-discourse around clouds is meteorological, that is, reading clouds for physical rather than metaphysical signs.  Farmers and sailors have known that so-called scud clouds are harbingers of rain, and Aristotle took a lesson about interpretation from the sky in the Rhetoric: “if it is cloudy, it will probably rain.”[1] Though it is difficult, as Friedrich Kittler points out, to separate weather and the gods, there have been efforts since antiquity to read the sky secularly and scientifically.[2] In The Nature of Things, Lucretius reproved people who found faces and animals in the clouds, saying that we should simply see clouds as the fortuitous motions of atoms.[3] Other ancient thinkers emphasized the random quality of cloud shapes and taught us instead to understand images in the clouds as figments of imagination, as resting entirely in the viewing subject and not in the nebular object, which only had physical data to communicate.  (In Calvin and Hobbes, Calvin spots a cloud and says, “it must be a sign!” Of what, asks Hobbes.  “Of very peculiar high altitude winds, I guess.”)  This debunking reading of clouds is anti-theological in the case of Lucretius and anti-philosophical in the case of the comic playwright Aristophanes, whose play, The Clouds, mocks Socrates, his head-in-the-clouds thoughts, and anyone who lives in Cloud Cuckoo Land.  In both cases, clouds are associated with airy, theoretical, insubstantial things, whether gods or “ideas.”

Even so, we should be careful, as there are elements in Jewish and Christian religion that are just as critical of reading God or anything else directly in the clouds.  The Prophets were fiercely iconoclastic and denounced the reading of clouds for portents or omens, and God stumps Job by asking him to explain the origin of clouds, as if this were something that would forever elude his understanding.  Jesus, squarely in this tradition, when asked to provide some curious spectators a sign in the heavens, gave an answer as sarcastic as anything in Lucretius.  If you want a sign in the sky, he said, here is one: red sky at night means good weather tomorrow, and red sky at morning means bad weather.  His “sign” was a lesson in everyday meteorology.[4] Those who say there are no gods in the sky or say there is only one can agree that it is foolish to look for divine signs in the clouds; the anti-divination discourse about clouds comes from both atheistic and theistic positions.  Yet in popular religiosity the clouds remain something to conjure with, habitations for all manner of heavenly entities.

The debunking discourse around clouds comes down especially hard on the practice of finding animal and human shapes in them.  Hamlet’s toying with Polonius must be the most famous example:

Hamlet: Do you see yonder cloud that’s almost in the shape of a camel?

Polonius: By th’ mass and ’tis, like a camel indeed.

Hamlet: Methinks it is like a weasel.

Polonius: It is backed like a weasel.

Hamlet: Or like a whale.

Polonius: Very like a whale. (Hamlet, III.ii.361-367).

Polonius’ foolish suggestibility, as mutable as the clouds themselves, enforces the idea that clouds completely lack objectivity, a notion found in the use of the cloud-like “thought bubble” in cartoons, a convention that links mental privacy with clouds.  In this discourse, clouds are the embodiment of all that is flighty, unstable or projected and thus the basis for comic commentary on how people manage to create meaning out of the blue.  Since anybody can find anything in them clouds stand for the unreliable treachery of perception; they stand for subjectivity itself and show how whimsical interpretation can be.  Thus some scientists call humanistic research “cloudy” or “sky-writing.”  Nonetheless, the British Cloud Appreciation Society has published a coffee table book and maintains a website dedicated to charming and droll images of “clouds that look like things”; humorous routines about what clouds look like remain a staple in popular culture.[5]

One of the standard jokes in the that-cloud-looks-like-a-? repertoire is “that cloud looks like a cloud.”  In fact, there is a strong tradition in the past five centuries of looking intently at and reading clouds, in art, media, and science.  Take art first.  Clouds have been a repeated and prominent subject in European painting since at least the Italian Renaissance.  The puzzle is why clouds would proliferate when they seem to defy the revolutionary technique of linear perspective, which so changed western painting, drawing, and architecture.  Clouds lack clear edges, morph rapidly, exist as much in color as in shape, and are not exactly objects like other objects.  This is the question art historian Hubert Damisch explores in his great book on cloud painting.[6] Damisch sees in their difficulty a chance for painters to both defy the strictures of perspective and show off their painterly skills.  Clouds are the “other” to linear perspective–abstract, aniconic, sheer image without likeness.  We cannot say of a cloud painter such as Jacob van Ruisdael that he sees camels, weasels, and whales in the sky: rather he rigorously documents the clouds themselves in all their visual glory.  Meteorologists even claim to be able to find in his paintings reliable historical witnesses of weather patterns.[7] In cloud painting artists depicted a curious kind of image, one that was neither symbol, icon, nor index, but rather atmosphere and process, like the act of painting itself.  (There are as many clouds as there are painters: they offer hospitality to a wide range of styles without ceasing to look like clouds.)

In painting clouds were harbingers of a new kind of image, an abstract one of flow and turbulence rather than symbolic representation.  They were among the first abstract objects to be depicted, and in this they are a critical early step in the history of recording media.  Kittler has famously argued that the acoustic and optical analog media of the nineteenth century caused a critical historical rupture: with photography, phonography, and cinema, the realm of the recordable expanded drastically to include non-intelligible objects, breaking writing’s historical hold as the only medium of storage for any form of art or intelligence.  “White noise” [Rauschen] was Kittler’s preferred term for this new class of recordables, though this is also perhaps what cloud painters had been depicting.  Writing could record the words “he sneezed” but could never preserve the complex motion or sound of sneezing, something that became routine in the later nineteenth century.  Sound, motion, flow, process all became recordable and thus subject to time-axis manipulation and to analysis.[8]

The scientific standing of clouds benefitted from these transformations.  As historians of science have noted, sound recording and motion pictures should be placed into a wider range of nineteenth-century scientific instruments with names ending in -scope and -graph that allowed observation and inscription of temporal processes of all kinds ranging from blood pressure and weather to noise and heat.  The “graphic method,” as French physiologist and proto-cinematographer Étienne-Jules Marey called it, followed innovations in mathematics and modeling such as Fourier equations that could represent fluid dynamics such as sound, heat, and atmospheric aerosols (i.e. clouds).  New methods brought new objects onto the scientific agenda: heat, noise, smoke, glaciers, clouds.  Indeed, the nineteenth century opened with the standard nomenclature for cloud types being announced by the British gentleman scientist Luke Howard in 1802.  Though rivals proposed alternatives, his Latin-based classification of stratus, cumulus, cirrus, and their varieties has held steady.[9] More challenging were international scientific efforts in the late nineteenth century to assemble a “cloud atlas” made up of photographic images illustrating the various types.  The particularities of the clouds defied the demand for standardization that an atlas requires.[10] Clouds are never quite capturable by language or thought; John Ruskin, the patron saint of the Cloud Appreciation Society, who wrote eloquently of cloud beauty in modern painting, said that a cloud is a mix of something and nothing.  This mixed ontological status was one of the things that so fascinated nineteenth-century writers and artists such as Goethe and Emerson, Shelley and Baudelaire, Constable, Turner and Monet.  The sea of faith might have been draining in the nineteenth century, but many still took comfort in the clouds.[11]

A history of cloud science in the last two centuries would also be a history of our media technologies.  Around the turn of the twentieth century, the Scottish physicist C. T. R. Wilson invented the cloud chamber.  Seeking at first to model cloud formation and water vapor, he hit upon an instrument that, thanks to its ability to detect and trace subatomic particles, played an essential role in particle physics in the first half of the twentieth century.  If clouds were once seen as cloaks of reality, in the cloud chamber they had become its revelators.

A similar path between obscuring and revealing was traveled with satellites and computers.  During the Cold War, high altitude spy planes could only take pictures of nuclear facilities and other targets on the ground if there was no cloud cover.  One impetus for weather forecasting was espionage: how to predict when clouds would part enough to justify a dangerous mission.  Clouds may have blocked the intelligence cameras but they revealed a great deal about weather to the meteorologists who learned to read their patterns.  Though people have long looked down on clouds from mountain tops, high-altitude aviation, space exploration, and satellites made possible remarkable new views of the earth, and one of the clearest facts about our planet is that it is covered with clouds.  Tracking and modeling clouds requires enormous amounts of data.  And though the standard story of modern computer science emphasizes the founding desire to model nuclear explosions and their aftermath, the demand for weather data has been just as formative in advancing digital technology.  (John von Neumann, for instance, was just as passionate about computer applications for weather as for nuclear explosions.)  The first world wide web infrastructure was arguably formed for watching weather, well before the web as we know it.  Clouds sit at the heart of crucial innovations.[12]

Fractal geometry was another spin-off of cloud study, which taught the logic of vagueness or the analysis of the indefinite.  Philosophers say clouds illustrate the “sorites paradox” or heap problem: two grains of rice do not make a heap, and you can remove two grains from a heap without it ceasing to be a heap, but by adding or subtracting either a heap will come or cease to be.  There is a vague boundary between heap and non-heap that can never be numerically specified.  In the same way, there are many possible surfaces that can plausibly claim to be the edge of a cloud.[13] This ontological indefiniteness is part of the cloud’s great intellectual fascination.

It is easy to say that clouds do not mean anything, but the deeper fact is that clouds mean a great deal and that the collective future of the human species may depend on reading them well, at least if we think about the ever-rising anthropogenic concentration of atmospheric carbon and the radical changes to climate it implies.  Now we face clouds that are no longer undisturbed natural artifacts.  Smoke stacks, nuclear bombs, cloud seeding, and geo-engineering schemes show that many kinds of the most important clouds are anthropogenic.  The artifactual character of clouds is emphasized by recent artists such as Fujiko Nakaya, who builds cloud and mist installations, Berndnaut Smilde, who creates and photographs clouds inside of buildings, or “Monsieur Moo” who, in a performance about the legal ownership of clouds, transported rain cloud balloons across the US-Canada border in violation of international law.  (Indeed, there are treaties about cloud-manipulation, e.g. causing flood or drought, for military purposes.)  Clouds are the exact sort of things that Bruno Latour likes to call “hybrids” or “imbroglios.”

The first result on a Google search for “cloud” I get is an ad for “Microsoft Cloud,” with an image of a data center topped by a puffy cumulus, as if Microsoft benefitted from a celestial benediction.  This use of the term “cloud” may have started in engineering diagrams of networks, but it almost instantly took to the sky, taking selective advantage of the surplus and residue of the term.  The web is full of beatific images of laptops sitting on heavenly clouds.  The rhetoric of the data cloud likes to exploit the peaceful, inconsequential parts of the tradition while suppressing the rest.  “The cloud” is a huge PR achievement for the IT industry, but it is profoundly deceptive.  For one thing, “the cloud” of online storage is neither natural nor environmentally friendly: It consists of a gigantic infrastructure of data centers.  For another, the notion of “the cloud” downplays the risk of giving up control over our data.  We might think about the term “cloud-attack” from World War I (a barrage of poison gas) or “cloud-burst” from meteorology to counter the blithe IT ideology of the cloud.  To trust our data to “the cloud” may invoke old ideas of the benevolent gods above, but the more interesting part of the history of clouds is how much human-built meaning is there to be exploited if you know how to do so.  The IT industry would like us to recall nothing but cloud illusions, as Joni Mitchell sang, but in this case, it is better to try to know clouds from both sides now.  In all moments of history, this would be the worst to think of clouds as purely immaterial, natural, and meaningless things.


1. Aristotle, Rhetoric, 1393a.

2. Friedrich Kittler, Musik und Mathematik 1:1 (Munich: Fink, 2006), 79.

3. De Rerum Natura, book 4, lines 166ff.

4. Matt. 16: 2-3.

5. For a large and diverse list of examples of cloud-gazing in popular culture see

6. Hubert Damisch, Théorie du nuage.  Pour une histoire de la peinture (Paris: Seuil, 1972).  Translated by Janet Lloyd as A Theory of /Cloud/ (Stanford: Stanford UP, 2002).

7. Franz Ossing, “Haarlem’s Crown of Clouds: Meteorology in the Paintings of Jacob van Ruisdael,” trans. Kari Odermann, (accessed 3 May 2014).

8. Gramophone, Film, Typewriter, trans. Geoffrey Winthrop-Young and Michael Wutz (1986; Stanford: Stanford UP, 1999).

9. See Richard Hamblyn’s highly readable but slightly hagiographic The Invention of Clouds: How an Amateur Meteorologist Forged the Language of the Skies (London: Picador, 2001).

10. Lorraine Daston, “The Science of Clouds, or: The Limits of Representation,” University of Oslo, 13 Sept. 2012.

11. On clouds in nineteenth-century thought and art, see Kurt Badt, Wolkenbilder und Wolkengedichte der Romantik (Berlin: de Gruyter, 1960); André Weber, Wolkenkodierungen bei Hugo, Baudelaire, und Maupassant im Spiegel des sich wandelnden Wissenshorizontes von der Aufklärung bis zur Chaostheorie (Berlin: Frank und Timme, 2012), and Wolken: Welt des Flüchtigen, ed. Tobias G. Natter and Franz Smola (Ostfildern: Hatje Cantz Verlag, 2013).

12. The whole paragraph owes to Paul N. Edwards’s excellent A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming (Cambridge: MIT Press, 2010). (accessed 8 May 2014).  

Comments are closed.