Strange Glyphs From Another Time

Question

The following glyphs were recently uncovered at a dig site, which have stumped cryptographers and linguists alike. These symbols are guessed to be some sort of ancient language, but nothing is yet known. Photos of these glyphs have been distributed to news sites, newspapers, and magazines in hopes that a member of the puzzling community might solve the enigma and uncover what this language says.

Studies on the strange inscription are ongoing, and more clues are likely to be uncovered as time goes on...

News Updates (Hints)

A linguist recently pointed out that the entire inscription is outlined, and that this likely denotes the inscription as a whole is meant to be grouped as one single thought, possibly conveying a single phrase or sentence rather than multiple.

Leading cryptanalysts have determined that the varations in line thickness and length create too wide a spread of possible patterns if they were to be considered as differentiators. Therefore, these variables are being discarded and assumed as inconsequential to the translation. A similar assumption has been made for the sizes of dots and circles, however while overall size is being ignored, differences in relative size of these objects are still being investigated.

It was reported yesterday that a bored archaeologist stumbled across a very strange connection between the uncovered inscription and a fictional television show during a work time web-surfing session. Quite oddly, this inscription has undeniable similarities to numerous fan created languages known as "Gallifreyan" and potential similarities between these language systems and the inscription are being investigated. Historians are left reeling at the possibility that this link is more than just purely coincidental.

Archaeologists have noted that this inscription was found on a circular tablet, without any additional markings aside from the glyphs in question. This peculiarity leaves no indication on which end should be up or down. It's been theorized by linguists that perhaps neither cardinal directions nor overall rotational orientation alter this language's ease of readability.

Cryptanalysts of multiple firms have come together to declare a standard terminology and categorization of symbols identified thus far, in the hopes that it would help all solvers no matter their corporate ties.

It has been found that there are 20 "groups" of circles, which are theorized to be words of some kind. These have been dubbed the "word circles".

All word circles appear to be connected by a spiraling line, which has been dubbed the "inscription spiral" for ease of reference, and may provide some indication of a starting point or translation direction.

Each word circle seems to comprise of one main encompassing circle, to which smaller circles are either inlaid, overlaid, and/or crosscut. These are being referred to as "subcircles".

Subcircles seem to have either one or two white dots, or "bubbles" nearby, which are similar in relative size to black dots, or "dots" for short.

It seems in all cases, dots and bubbles are clustered in close proximity to subcircles and as such these are theorized to be modifiers to the subcircle.

Aside from the inscription spiral, it seems that lines fall into two categories: "orbital lines", which orbit a subcircle by following its curvature, and "non-orbital lines", which seem to be much less restricted and are drawn almost freely. Furthermore, non-orbital lines appear to always connect to a bubble on at least one end.

Certain subcircles appear to have a single unbroken circle "inlining" it. Interestingly, it appears that this "inline circle" only appears on word circles which have more than one subcircle.

The double outline around the entire enscription has been dubbed the "inscription outline". Given the assumption that every word circle is in fact a word, leaving 20 words in the inscripton, linguists have theorized that the outline is in fact the punctuation mark for what is likely a sentence.

An interesting observation has been observed by a student intern at a leading cryptanalysis firm. It seems that every subcircle modifier with dots can be categorized into two buckets: dots which are clearly and obviously attached to one subcircle, and dots which are not clear to which subcircle they modify. However, all dots which are not clearly attached to one subcircle seem to be equidistant between two subcircles in every case. The theory is that these dots count for both subcircles that they are equidistant between, thus eliminating the guesswork to which subcircle a dot applies. If correct, this property likely applies to shared lines between bubbles as well. Rumor is that the new-hire paperwork for this intern is already underway.

Archeologists have hit on a major discovery today, as a second inscription has been uncovered. Astoundingly, this tablet has a phrase written in Latin across the back, which translates to "A quick brown fox jumps over the lazy dog!" If archeologists are correct, this means both the newfound inscription and the latin phrase mean the same thing, and could be a major resource going forward. In light of this, CEO of the cryptanalysis firm leading the decyphering effort has ordered all holiday plans be canceled, to eliminate any unwanted distractions while this new discovery is analyzed.

At this point, the question was solved. However I am adding the last two clues I had prewritten just in case there happens to be a late solver in the future.

A world renowned linguist recently published a book titled "Improving The World Through Phonetically Consistent Languages", describing the societal benefits of a world where cacoepy is nonexistant. While critics have dismissed the work as "just some linquist's wet dream", it has left some forward thinkers wondering what such a language might look like...

Self proclaimed "cryptographic genius" has made an announcement on internet forums that he has identified a critical clue regarding subcircle modifiers. While there seems to be no up or down to the inscription (due to the inscription being found engraved on a circular tablet), this does not mean that there is a complete lack of direction. He points out that inwards and outwards are still valid directions when it comes to circles, and that this may be important for understanding ongoing confusion regarding bubbles.

Solution

The following is a full explanation and solution key for the cipher. Obviously, big spoilers follow.

When I created this form of Gallifreyan, I started with Sherman's Gallifreyan, as it was the first form of circular gallifreyan and is the most widely known form available (there are many more). While I found the language to be extremely creative and beautiful, there were two things that did not quite sit right with me.
1) If this is truly a language, why is it just an encoding of English words, using the English alphabet and language?
2) Circles have no absolute directions, so if this was a truly circular language, why would there need to be an indication of which side is "down"?

Fixing these items was the basis of my new version of Gallifreyan, which I call Omnidirectional Gallifreyan, and it is what lead to the creation of the cipher I posted here. I first set about creating a modified version of Sherman's, taking inspiration from Doctor's Cot where each symbol was based upon a phonetic sound. Secondly I created rules which adhered to a circular theme, ensuring the only directions needed were inward, outward, and counter-clockwise, this way the start of a word did not have to depend on the orientation of the glyph.

The following is the key I used to encode the sounds into a written form. While this key only includes the sounds used within the English language (because I am an English speaker), it could be expanded to include every phonetic in the IPA, thus creating a complete, language agnostic form of Gallifreyan... I may do this in the future.
An important thing to note with vowels: Vowel bubbles are always either on the line of the consonant circle, or within the consonant circle and not intersecting. The upper intersecting bubble in this chart represents a bubble that faces the center of the word circle, while the lower intersecting bubble represents one which is facing away from the center of the word circle. The denotation whether it is an "inwards" or "outwards" facing bubble is whether it is inside or outside of the intersection between the word circle line and the consonant circle line. For vowels written on a consonant circle that is completely encompassed by a word circle, it should be obvious whether that bubble is closer to the edge of the word circle, or the center. For example:

Word Structure

The starting phonetic of a word must be "inlined" by a full, unbroken circle which is close in size to the consonant circle. An exception to this is if the word circle only contains one consonant circle, in which case the inline circle may be omitted.

Phonetics are pronounced in couter-clockwise order beginning with the starting phonetic. Consonant circles may contain at most, one vowel bubble. If there is a need for sequential vowel bubbles, the silent consonant is used to separate them.

Consonant circles come in four types as seen on the chart, and the upwards direction represents inwards towards the circle. Consonants are modified by either dots or moons, which is what I call the orbiting lines. Vowel bubbles come in four types as well: Inwards, outwards, within, and double within. Note that the double within bubbles do not need to touch (I did this in my cipher simply for consistency reasons, to make decryption a bit easier). These vowel bubbles are modified by 0-3 rays. It does not matter what vowel rays connect to, only how many endpoints are attached to that vowel bubble. A vowel ray that circles around and connects to the same vowel is considered two endpoints.

Sentence Structure
Given that this is a phonetic language, there are no grammar rules. No commas, hyphens, or weird semicolons to deal with. At most, a silent consonant without an attached vowel may be used to indicate a pause in speech similar to a comma, and a chain of them used to denote a longer pause like ... does. This is the extent of spoken language grammar.

While I do not have a hard rule for how to denote the start of a sentence, it is recommended for it to be clear to the reader. Ideas for how the sentence start can be made obvious, is to make the first word circle jut out from the rest of the inscription, or to make the first word bolder.

Starting from the first word, subsequent words are denoted by a line connecting one word circle to the next. This "sentence line" must start and end on either a word circle, or a consonant circle. It may not attach to bubbles, dots, or moons. For artistic purposes, this sentence line may be omitted if the word circles form a clear direction of flow, such that each word circle is only close to one or two others (the previous and/or subsequent words).
If it is desired, the inline circle designating the starting phonetic may be omitted in words which are not the start of a sentence. In these cases, the starting phonetic of the next word is the consonant circle that the sentence line connects to, or the next consonant circle counter-clockwise from where it intersects with the word circle. This also helps identify the starting word in a sentence.

Completed sentences should be outlined to denote it as one complete thought. The number of outlines defines whether that thought is a question, a statement, or an exclamation. This sentence outline functions the same as punctuation in our languages.


Multiple sentences are connected by overlapping sentence lines. The subsequent sentence is outlined by the sentence line that goes beneath the previous one.

A visual example putting everything together:

That's it! Now you can write in Omnidirectional Gallifreyan too!

For the original cipher inscription, the following is the phonetic encoding I used. I have bolded the consonants to make them easily distinguishable from vowels.
A straight line may be the shortest distance between two points, but it is by no means the most interesting.
אeɪ streɪt Ɫaɪn meɪ bi ðʌ ʃoʊrtɪst dɪstɛns bitwin tu pɔɪnts bʌt אɪt אɪs baɪ noʊ minz ðʌ moʊst אɪntrɛstɪŋ.
Note that I made a small error during cipher creation, and encoded is as אɪs instead of אɪz. Oops.

Answer 1

There is a ton here to unpack and it is a lot to explain an entire alien language in a post.

This is, as noted by many, a version of modified Circular Gallifreyan, perhaps a known one or a new one, I’m not sure. It doesn’t match the Loren Sherman guide at least. https://shermansplanet.com/gallifreyan/guide.pdf

In Gallifreyan, circles represent words. Circles contain smaller circles which represent consonant sounds (not letters; C in cat and K in kitten would be the same symbol). Even smaller circles represent vowels sounds (again, O in two and Ue in Flue would be the same symbol). The sounds are read counter-clockwise; starting point is the circle with double-bordered circle, when applicable (credit Christoph). The full text reads “a straight line may be the shortest distance between two points, but it is by no means the most interesting”; which is a quote from Doctor Who, the show where Gallifreyan is spoken, so quite on theme.

I think the easiest way to understand this is visually. In the marked up image, the patterns become pretty clear, and near the end of this post, Deusovi has an even clearer version with the words annotated.

There is an overall spiral line which marks the sentence and does not impact letter meaning (starts on left and spirals inwards). I had tried putting together a chart explaining each letter, but got stumped on a few vowels and 'S'. At the end of this post, Deusovi has kindly created one, it seems to match my findings. Some examples:
Long A (ay) is a small open circle with 2 lines.
T is a semi-circle with a black dot.
‘N’ is a full circle, except it is accompanied by a black dot. Sometimes the dot is ‘inside’ the N, sometimes it is nearby; both are valid, as in normal Gallifreyan.
And so forth...

Here's a full diagram of the sentence:

(This uses the IPA for phonetic notation. Most of the consonants are what you'd expect; vowels are always pronounced as in Spanish or Italian. ð is "mother", ʃ is "ship", ŋ is "sing", ɪ is "sit" (as opposed to i, "seat".))

A few symbols are circled in red - those appear to be mistakes.

The assignment used is shown here:


Interestingly, there are some nice phonetic patterns, particularly in the top left (m/n/ŋ above p/t/k/b/d). But those patterns don't continue to the rest of the grid.

Answer 2

Ok, this has been here for a while with no answer, so I'm going to document my (limited) progress here so that others might use it as a stepping stone, or point out the flaws and blind spots in my thinking. It might also give @ExecutionByFork an idea of where future hints might help.

First, what is this?

As noted in comments and since confirmed in a hint, this is a variety of Circular Gallifreyan. Assuming something similar to Sherman's Gallifreyan, each larger circle represents a word, and the smaller circles+dots/lines around them represent letters/syllables/phonemes(ish).

But you can get all that from a few minutes Googling, so what next?

It looks like there are 20 words, with a spiral path indicated by the line behind them, so we probably want to follow that line and read clockwise going inward, or anticlockwise coming out. I've assumed clockwise going in, because the small word on the left is more like to start a sentence than end it.

Now, the letters themselves are clearly quite similar to Sherman's, but not a direct copy, so what do we have?

There are five circle styles, the first four of which appear in Sherman's, the fifth is new here:
- 3/4 cutout
- Inside circle
- 1/2 cutout
- Edge Circle
- Semicircle (new!!)

Some of these appear with an extra circle (doubled?)

There are then a bunch of other ornaments:
- white dots - in various positions, with varying numbers of additional lines
- black dots - in various positions and numbers
- additional circular arcs tangent to the letter circle

Note that we never see black dots and circular arcs at the same time, which makes them look a lot like Sherman's consonant modifiers. But we clearly aren't using the same encoding.

And that would lead me to guess that the white dots are similar to Sherman's vowels, but with more variety in position and line count.

So how do we interpret/translate it?

If we could be sure that we are interpreting the symbols correctly, then we could assign them arbitrary letters, and treat the resulting string as a simple substitution cipher.

But even if all the assumptions above are correct, we still have a lot of variables to play with. And all of these uncertainties multiply up to an awful lot of cases to check.
- There are too many different vowel symbols. The single white dots appear on the edge to one side or in the middle of the letter circle, with 0-3 lines going in various directions. Even if we simplify this down to middle/edge and ignore line direction, we have seven distinct vowel modifiers. (Single dot with 0/2/3 lines, and the double-dot with single line.)
- The black dots appear inside, outside and on the edge of the circles. For now I'm assuming that only the count is significant, but this could easily be wrong.
- There are several black dots that could belong to more than one letter, so lots of potential for mistranslation when you multiply up all the possible ways to misread them.
- I'm assuming that the black dot between words 16 and 17 (counting inwards) is punctuation, but it could belong to either of the adjacent letters.
- Do the "doubled" circles mean double letters or something else?
- Does this uses digraphs (as Sherman's does)? If so, it adds another layer of uncertainty to any attempted transliteration.

And that's before I consider what order to read the letters in (so far I've assumed anti-clockwise starting from the bottom (as for Sherman's)), but that is just another assumption on the house of cards I've built so far.

The following is unspoilered because tables in spoiler blocks are a pain, and it's probably wildy wrong anyway. But it may be useful to someone else.

Applying the assumptions above, the following tables gives the count of each symbol, and the transliteration I have used for attempted decryption.
For consonants, I'm basically starting with Sherman's assignment and rearranging a bit to fill gaps and avoid digraphs.

		1 Dot	2 Dot	3 Dot	1 Arc	2 Arc
3/4 cutout	1-B	12-X		2-D		1-H
Double 3/4 cutout	1-BB		1-QQ	1-DD	1-GG
Inside circle		2-V			7-N
Double inside circle					1-NN	1-PP
1/2 cutout	1-T
Double 1/2 cutout						1-WW
Edge Circle	4-F	6-M		2-Z
Double edge Circle	2-FF
Semicircle	1-C
Double semicircle	3-CC

For vowels (white dots), this is arbitrary, with a couple of consonants filling in because I don't have enough vowels to go round:

	0 line	1 line	2 line	3 line
Edge	2-A	-	3-E	2-I
Centre	4-O	-	2-U	6-Y
Double	-	6-S	-	-

Using those gives the following transliteration (reading words clockwise/inward from the left, and letters anti-clockwise from the bottom of each word):

1	CE
2	ME	X	NN	X
3	WWI	M
4	FE
5	DO
6	VY
7	XS	N	X	PPY	Z
8	GGS	N	XA	M
9	M	QQO	X	TO
10	XU
11	BBU	M	X
12	DD	Y	B
13	CCS	X
14	CCS	N
15	DI
16	FY
17	H	FFO	F
18	VY
19	X	FFY
20	F	X	ZA	N	XS	M	CCS

Hopefully some of that is useful, if not I'll be deleting it in a few days.