Hieroglyphs

The work and its context

Hieroglyphs asks how little geometry is needed before marks begin to look like signs. A diagonal, a corner, a triangle and a square can appear as separate characters, yet a change in spacing can make them merge into a continuous pattern. The four works examined here move between those conditions. Their difference concerns not only the marks themselves, but whether the viewer encounters a sequence of units or an uninterrupted field. ¹²³⁴

Pxlq released the 1,000-work project through Art Blocks Playground in March 2021. It is project 30 on the platform's second Ethereum contract. The official description presents it as an exploration of abstract symbols through simple rules. The pxlq dossier provides context for the artist's broader work, while this essay concentrates on the specific drawing procedure and its visible consequences. ⁵

The title encourages a reading associated with writing, but the inspected outputs offer no key to a language. In one, separated tan shapes resemble an inventory of characters. In another, repeated yellow strokes form a regular lattice. The invitation to read remains useful even when no message can be recovered: attention moves across the surface, searching for recurrence, difference and grouping. This is a critical interpretation of the works, not a documented claim that they encode a historical script. ²⁴⁶

Conception and development

The available primary record is concise. The collection description identifies generative symbols as the project's subject, but supplies no dated prototype sequence, development diary or account of a particular historical influence. The inspected code establishes how the deployed system works. It does not establish how long pxlq spent making it or which discarded experiments preceded it. Those parts of the project's conception remain undocumented here. ⁵⁶

The release sits within an existing Art Blocks practice. The platform dates Dynamic Slices to 12 December 2020 and Cyber Cities to 7 February 2021, before Hieroglyphs on 21 March. Their respective classifications are Curated, Playground and Playground. This chronology establishes successive releases by the same artist; it does not prove that the earlier programs supplied components of this one. A claim about code reuse would require comparison of the relevant versions. ⁷⁸⁵

What can be assessed directly is the economy of the final system. Instead of requiring a large catalogue of predesigned signs, it generates paths within a cell and controls how those paths repeat. The four examples show the consequence of that choice: recognizable elementary forms can produce either a dispersed symbolic texture or a nearly continuous composition. This account of the design follows the program and the viewed works, rather than filling the absent development history with a plausible narrative. ⁶¹²³⁴

How the work was constructed

The indexed program is unusually compact. Its principal unit is a square cell whose corners can be selected by pairs of binary values. A pair chooses left or right for the horizontal coordinate and top or bottom for the vertical coordinate. Sequences of those choices become paths, and two paths are assigned to each cell. The apparent vocabulary of signs therefore begins with four available positions, rather than a stored alphabet of finished characters. ⁶

This distinction limits what the title can establish. The inspected program contains no dictionary of Egyptian signs, no language model and no mapping from a text to its visual equivalent. Its symbols are generated from geometric decisions. A reader may approach the result as writing, but the code does not provide a message to decipher. That is a formal inference from the procedure, not evidence of the artist's intention to represent a particular writing system. ⁶

A number derived from the beginning of the token hash seeds a small bitwise pseudorandom routine. Its successive results choose a palette, grid dimensions and other parameters. The palette list contains seventeen sets of three colours, used for the foreground, ground and optional offset layer. Selection is repeatable through the program's own numeric sequence; no external image or font supplies the marks. This source reading does not establish equal probabilities for every palette or visible configuration. ⁶

The grid count is chosen between three and twenty. A border surrounds the working field, while a separate branch either inserts spacing between cells or sets that spacing to zero. Cell size is calculated from the remaining room. Stroke widths scale with both the working dimensions and the count, keeping the weight of marks related to the grid rather than fixing it at one screen-pixel value. The distinction between separated signs and a continuous pattern consequently has a direct procedural basis. ⁶

Two further decisions determine how far a chosen pair of paths persists. A single pair may serve throughout the grid; it may be replaced as a new row begins; or it may be replaced for individual cells. These branches define different degrees of repetition. Repetition is therefore not simply an accidental resemblance among independently generated signs. In some configurations the program deliberately reuses the same instructions at many positions. ⁶

The routine that makes each path also needs careful description. Its loop condition draws a new random upper bound on each test. It does not select one fixed path length and then fill exactly that many positions. A description of uniformly selected lengths would therefore misstate the inspected code. Likewise, the drawing routine closes the shape after each added vertex, rather than only after the entire sequence. In p5.js 1.0.0, successive endShape calls retain the vertex list until the next beginShape call. This makes the intermediate stages relevant to rendering, especially where compositing is enabled. ⁶⁹¹⁰

The optional shadow is another drawing pass, offset by five percent of the cell's width and height. All such offset marks are drawn before the foreground pass. It is a flat translation with a separate colour and narrower stroke, not a calculation of lighting or an object's volume. One palette even assigns a transparent value to this layer. An additional branch enables hard-light compositing for a subset of the smaller grids. Those choices change the relationships between overlapping marks without adding more kinds of geometry. ⁶

Rotation acts at the scale of the whole field. When selected, the working dimensions exceed the visible square and the drawing is translated and turned by forty-five degrees. The canvas then shows a cropped portion of that larger arrangement. A rotated result should not be described as a square grid merely spun in place: the enlargement and translation are part of how the visible field is composed. ⁶

All drawing occurs in setup. There is no artwork draw routine, animation clock or interaction handler in the inspected script. Its normal result is static, although the precise appearance still depends on the pinned drawing library and browser rendering. The four canonical renderings discussed below were checked directly; no cross-browser pixel comparison was performed. ⁶

The mint

The official collection page gives 21 March 2021 as the release date. The index records activation at 21:49:35 on 20 March and the first token at 22:18:04 that day. The second mint and scheduled start are recorded at 17:00:00 on 21 March. Completion is recorded much later, at 16:47:15 on 26 June. Full timestamps and offsets are retained in the evidence record. These are distinct indexed events; the earlier first token does not by itself establish a public release on 20 March. ⁵¹²

The observed invocation count and maximum are both 1,000, with a current price setting of 0.1 ETH. That setting is not verified here as the original opening price or the amount paid by every recipient. Nor has the division between reserved, artist and public mints been reconstructed. The months separating the recorded start and completion make a rapid-sellout story particularly inappropriate, but they do not establish a complete account of availability or configuration changes during that interval. ⁵

The sampled token records retain transaction identifiers as evidence for further investigation. They anchor the examples to specific works without resolving the missing edition-wide purchase history. No precise public-sale duration, aggregate proceeds or gas expenditure is asserted. ¹²³⁴

Copyright and licensing

The observed artwork label is NFT License. The index supplies no terms link, and no operative text or version was established in the inspected sources. This label is recorded as unresolved rather than interpreted as unrestricted permission. A history of earlier artwork terms also remains to be recovered. The collection-image caption identifies this reported status beside the official token link. ⁵

No separate reuse grant for the artist's complete program was found in the inspected script. The p5.js 1.0.0 dependency has its own LGPL version 2.1 licence, available in the tagged repository. The implementation files consulted to understand shape drawing belong to that library record. Their licence cannot simply be transferred to pxlq's program or to the resulting artworks. ⁶⁹¹¹¹⁰

The original essay prose is dedicated to CC0. The full artist script remains research material rather than a redistributed component of this publication, and its hash identifies the version inspected. Library rights, artwork terms and artist-code permissions remain distinct parts of the record. ⁶¹¹

Close reading of four outputs

Editions 0, 1, 10 and 20 were selected in advance as a small early-edition comparison. Their full token identifiers are 30000000, 30000001, 30000010 and 30000020 on Ethereum contract 0xa7d8d9ef8d8ce8992df33d8b8cf4aebabd5bd270. Each canonical renderer was viewed as a square artwork within a wide browser viewport. This is direct inspection of four examples, not a survey of the full edition or an audited feature distribution. ¹²³⁴

In #0, large cream polygons occupy an orange field. Dark offset edges give the flat shapes a slight impression of being raised from the ground. Diamonds and broad diagonals dominate, with several forms cut by the canvas boundary. The scale makes it difficult to sustain a reading of individual characters. Instead, the image presents a handful of competing planes whose edges direct attention across the square. Cropping makes the arrangement feel larger than the visible field. ¹

In #1, tan forms sit in a regular grid on very dark brown. Solid squares, triangles and notched polygons alternate with open corners and thin diagonal strokes. Gaps keep the units distinct, allowing the eye to proceed from one to another as if comparing entries in an alphabet. The uneven balance between filled and linear forms gives the grid its rhythm. A solid square arrests attention differently from a single slash, even when both occupy comparable cells. ²

In #10, very dark navy and pale grey create a more continuous field. Neighbouring filled regions appear to join, while diagonal cuts and outlined polygons divide them again. Some pale areas seem enclosed inside a dark mass; elsewhere they open onto the surrounding ground. Isolated strokes at the perimeter make the underlying cell logic easier to notice. Compared with #1, the work puts greater pressure on figure and ground: a shape can be read either as a mark or as the space left between marks. ³

In #20, fine yellow vertical and downward diagonal lines repeat across blue. Their consistency produces a dense lattice with a clear surrounding margin. Dark offset marks introduce a slight separation between the bright lines and the field, but the overall effect remains spare. Unlike the varied signs of #1, recurrence dominates almost every part of the square. The work demonstrates that a system associated with generated symbols can also produce a nearly uniform pattern, changing the viewing task from identifying differences to following continuity. ⁴

Together, these outputs show why spacing and scale deserve as much attention as the inventory of shapes. A diamond in a sparse enlarged composition does not have the same role as a small filled unit in a dense grid. Likewise, repeated diagonal strokes can act as independent signs or cooperate in a lattice. The code's limited vocabulary is sufficient because its rules also determine the relationships between units. ⁶¹²³⁴

Significance and limitations

Hieroglyphs makes the boundary between symbol and pattern available for comparison. It does not need a different drawing language for every output. A small set of positions, repeated paths and changes in spacing can redirect the viewer from character-like units to large planes or fine networks. The four examples reward attention to that change in organisation rather than to a presumed hidden message. ⁶¹²³⁴

Its place in early Art Blocks is therefore more specific than a general claim that code can produce variety. The project shows how much perceptual difference can arise when the identity of a mark depends on its neighbours and on the scale at which it is encountered. The available evidence supports this formal account more strongly than a detailed story of the artist's development process. Keeping that distinction visible leaves room for further research without weakening what can already be observed. ⁵⁶

Unresolved research questions:

• Locate a first-person development account, dated prototypes or launch-era documentation specific to Hieroglyphs.
• Recover original price and edition announcements, configuration changes and mint-allocation classifications.
• Establish operative artwork terms and any separate artist-code licence.
• Compare the indexed script with a pinned chain retrieval and test rendering across supported environments.
• Extend visual inspection beyond four early editions without treating a small sample as representative.