Honskhardn: Done!

Finally finished up the whole grammar.

Google doc here.

Progress lagged quite a big at the end, when I started needing to come up with text examples. This grammar won’t have a major text example as Pretistelen grammar did, largely because I just couldn’t think of anything to write for it.

Probably going to be a lot of updates to it later, since I’m still working on the Pretistelen lexical cognates in all the daughter languages. Indeed, I’ve already changed how two particular words in Pretistelen work (which means I’ll have to be changing the Pretistelen grammar), and that’s gonna create all sorts of chaos.

A lot of the text examples are probably a bit sloppy, since I last minute changed major elements of word order — in particular changing the prepositions back to postpositions (as the adpositions are in Pretistelen) and the order of the object and the verb from VO to OV.

 

Honskhardn: Pronouns

I’ve been having a little trouble working out the pronouns for this project.

Pretistelen, the origin language, has three personal pronouns of relevance:

Na, no, ne, né, nu, ni: I/we

Ta, to, te, té, tu, ti: You

Sa, so, se, sé, su, si: He/she/it/they

Each of these inflect for gender and number (a/é = feminine singular/plural, o/u = masculine singular/plural, e/i = neuter singular/plural).

Since monosyllablic words aren’t affected by the final-syllable dropping, these remain more or less unchanged into Hónskhardn.

However, I did adapt the 3rd person pronoun (sa, so, &c) to also function as a definite pronoun, to carry a word’s gender and number when, due to this syllable-dropping, it’s no longer clear on the word itself (kinda like how German uses its pronouns).

(So nid “fish” can become either sae nid “the (one) fish” or si nid “the fishes”, depending on what’s being talked about.)

Pretistelen is supposed to be spoken in a relatively developed society, and it’s become my understanding that developed civilizations tend to start doing weird things with their pronouns, like develop formality levels (see: late Latin, Chinese, Japanese, Hindi, various Mesoamerican languages, &c), which generally seems to corrolate with heavily stratified societies.

The easiest way for me to do this was with a pair of suffixes I’d developed for nouns and adjectives: -ól-, the diminutive suffix (nid “fish” > nidól “tiny fish”), and -aed-, the magnificative suffix (nid nidaed “huge fish”).

The general plan has been to create a three-tiered pronominal formality system for all the daughter languages of Pretistelen, applied to first- and second-person pronouns.

The first of these would be the informal/neutral set, which would be the bare pronouns (i.e. na, nota, to, &c), to be used between friends, family, co-workers, and people of similar social standing.

The second of these would be the formal set, modified with the magnificative suffix (naed, taed — note that the gender/number markers are dropped), the first-person used by a superior addressing an inferior, the second-person used by an inferior addressing a superior.

The third of these would be what I’m calling the “differential” set, modified with the diminutive suffix (nól, tól), the first-person used by an inferior addressing a superior, the second-person used by a superior addressing an inferior.

On the other hand, this doesn’t seem to be how formality systems evolve in natural languages? Which honestly kinda puzzles me, cause this system feels pretty logical, but probably a potentially pejorative pronoun wouldn’t really gain traction, save in certain contexts (such as between enemies).

Alternatively, with Hónskhardn at least, I’ve considered going a more European route, with the common-gender (that is, the feminine and masculine) plurals becoming numberless formal pronouns, and the neuter gender plural becoming the standard plural pronoun. (NOTE: the Neuter in Hónskhardn works differently than in English, as it’s not supposed to carry connotations of inanimacy; it’s more of an epicene.)

Thoughts?

Honskhardn 3: Verbs

So last time we did nouns, now we’re doing verbs.

So Pretistelen had verbs that agreed with the subject noun in person, number, and gender, in addition to having inflections for time, aspect, mood, and voice.

Cue Hónkhardn messing with everything cause it’s dropping the vowel of a word’s last syllable. (You may notice this has been a trend.)

Or rather, messing with the agreement markers.

Each of these elements are very specifically placed on the verb.

[Aspect] — VERB STEM — [time] — [mood] — [(gender/number agreement)] — [voice/person agreement]

EX: hokin-hokin-toll-j-ý-n “We will have to be cooking.”

Aspect, at least, has been totally unaffected. I may end up adding more aspects, perhaps, or deleting the two-way distinction that has existed so far, depending on if I see any verbs that could become satisfactory helping verbs or particles or inflections. Or any other words that could do those things. So far, though I just have the imperfective-perfective dichotomy, with imperfective marked by reduplicating the verb stem.

Time was similarly relatively unaffected, though I find the original past marker -u- / -ut- and future marker -at- to be needlessly similar. So I replaced the future marker with -toll-, from the word for “down” (the language having a vertical temporal axis, rather than horizontal as in English).

I have two voices, active and passive. For some reason unknown to my present self, I decided to mark person and voice with the same suffixes in Pretistelen (and, because language is messy and does all sorts of crazy things all the time, decided to just whatever keep it). However, as the passive voice markers all contain a full syllable, this means that the subject agreement marker is untouched, unelided, on passive verbs.

EX:
ot (second-person masculine singular active) > –ts
ondhe (second-person masculine singular passive) > –ondh

(Nouns and verbs share gender/number markers.)

Which means that passive verbs inflect for gender and number, whereas active verbs don’t.

Active markers:
1st: –n
2nd: -ts
3rd: -s

Passive markers:
1st: ng
2nd: -ndh
3rd: -z

And then the mood markers. Oh boy. This is where things got really interesting. I have three moods, in addition to the unmarked indicative mood: subjunctive, necessitive, and imperative (the imperative being the Pretistelen optative).

Although, if I think about it, I may have a better way to doing the mood than I had previously been doing it.

The mood markers in Pretistelen are as follows: -wé- (sub), –re- (opt), and -jy- (nec). These were generally placed right before the agreement markers, and because adjacent vowels contracted, this ought to result in horrible mutant inflections, which would only stick around because they’d be stressed and thus resist any elision (since, technically, the elision happens on any vowel after the stressed syllable).

However, I feel like it’d be simpler to just cut the vowel part of the mood markers out, to reduce them to -w-r-, and -j-, so that the conjugations turn out simpler, rather than dealing with four different agreement systems. (Which, I understand is a thing languages do, such as ancient Greek, but is not something I’d necessarily like to deal with.) But I guess this is one of the things I’d like some thoughts on.

Finally, I have two non-finite forms (well, four, but they split into active and passive).

First is the participle, which is formed with the suffix -is (active) or -az (passive).

Then is the gerund/infinitive. Because the Pretistelen infinitive was only marked with a diphthong on the stem, it basically has no right to exist in Hónskhardn. Instead, I have had its role taken up by the gerund, which is marked with just -s (active and passive). Which I understand is the same as the third-person active marker, but German is a thing.

Thoughts? Questions? Lay ’em on me. I’ll be here all week. And hopefully I can get the whole thing finished in four days. I haven’t even started on syntax yet.

Honskhardn 2: Nouns

So I was planning to post this stuff more often. Whoops.

Oh well. Time for nouns.

I have four declensions, with six cases, two numbers, and three genders. I’m going to be talking about all three, not necessarily in order.

Of the daughter languages I created for Pretistelen, this one probably has one of the more divergent morphology compared to the others, since it deletes or simplifies all the word-final syllables (that aren’t stressed).

Cases

Pretistelen has five cases, which I expanded into six in Hónskhardn, mostly because of some of the peculiarities of the aforementioned word-final elision and because of how possessives function.

The nominative is the zero case, with no special marking, used for the subject of a sentence.

The genitive is marked with n, and is used for possessives and prepositions of originitve motion. However, possessives have additional suffixes after the genitive marker, so that the word agrees in person, number, and case with the possessed word. (EX. ro “fang” > rono garo “the puma’s fang”.)

The instrumental is marked with t, and is used for instrumental functions (“with” or “by means of”) and prepositions of station.

The instrumental originally also had a dative function, but the preposition ros “toward” ended up fusing with the instruemental suffix to form a separate dative case, marked with –tros. This solely covers indirect objects.

Finally, the accusative case is marked with m, and is used for the object of a sentence and for prepositions of approaching motion.

With all that in mind, for Hónskhardn I decided to split the genitive into a genitive and possessive, particularly because many nouns end up retaining their word-final vowel morpheme in possessives (and in the dative). (EX. sisa “water” > sis (nom), sisn (pos), sisan (gen), sists (inst), sisatrs (dat), sism (acc).)

However, I’m unsure if I should unload the prepositional uses of the instrumental onto the genitive, since all it’s being used for now is prepositions. And if I do that, should I also pull all the prepositions that require the accusative onto the genitive as well?

Gender and Number

I put these two together because in Pretistelen they’re all conflated together. Pretistelen had two numbers and three genders, and each combination had a unique vowel assigned to it.

Masculine singular: o
Masculine plural: u
Feminine singular: a
Feminine plural: é
Neuter singular: e
Neuter plural: i

However, due to the final-syllable elision, these vowels only appear in monosyllabic words and in the possessive and dative cases (granted, they also only appeared on nouns in one declension).

Gender assignment aside, the most interesting development from this has been that number is no longer marked.

While I solved this problem in other daughter languages by applying a gender-neutral plural suffix from another declension more broadly across the nouns, I kinda like the idea of having at least a few of these daughter languages losing plural marking on the noun itself?

That being said, unlike some languages around the world, plural marking is always obligatory, so I figured that the speakers of this language would attempt to create some way of continuing to mark number (and gender).

The third-person pronoun is a monosyllable and inflects just like an adjective, which inflects basically like a standard noun, so I basically turned it into a definite article (saesaso). Sort of like how the German definite article carries more of a noun’s grammatical meaning than the noun itself.

Declensions

Pretistelen has four declensions, which I’ve called 1st Declension, 2nd Declension, 3rd Declension, and 4th Declension.

Or, if you prefer, based on the letter each ends on, V-Declension, N-Declension, S-Declension, and M-Declension.

So far I’ve been talking just about the V-Declension (“V” for “vowel”), since it both covers the most number of nouns and it’s affected most by the phonological changes between Pretistelen and Hónskhardn.

The other three declensions each have their own dedicated plural marker, but don’t mark gender. I’ve anticipated, however, that a number of nouns in these latter three would be brought into the 1st Declension, since a number of these words have endings similar to the now vowelless V-Declension. Though, upon reflection, there’d also be some flow the other way, too.

Essentially, I’ve been putting any noun with a long vowel or a syllabic consonant in the final syllable into one of the latter three declensions, and putting all the others into the 1st Declension.

Which still manages to be tricky, cause many of the words with long vowels could work as 1st Declension nouns and some words without might maybe retain their original designation? (EX 1. ansuan “pearl” > ansún (nom), ansýn (gen-2) or ansúnn (gen-1)? EX 2. tircas “bird” > terks (nom), terkizn (gen-3) or terksn (gen-1)?)

Thoughts?

I’m thinking I should maybe supply more vocab for examples and so that people don’t get lost, but I’m not sure if a small list at the beginning is helpful, or just bringing up an example or two per point.

Introducing: Honskhardn (and Phonetics)

So apparently a few people were actually interested in this constructed language I’m working on this month.

Which is kinda cool but also a little frustrating, since this particular language is vested in a lot of prior work I’ve done — which is to say, there’s a lot inform people about before getting to the language itself.

So let’s get started.

So a few months back, I finalized a grammar for a constructed language I’d been working on for some time, called Pretistelen (< link to the Google doc). Pretistelen is meant to be a large piece of background worldbuilding for a set of novels I’ve been working on (incidentally, the same world that the Biocrystal system will be used in).

However, Pretistelen itself as described in the grammer I wrote is spoken 800 years before the events of the novel. So I’ve also been working on a set of daughter languages used in the novel’s temporal setting. In particular, four I’ve determined to be relevant to the story (out of 10+ I could have easily gone for, and six that I’d originally planned for).

The language I’m working on this month, Hónskhardn, is one of the less relevant of the four, and the least complicated, historically and culturally; so it felt like the best daughter language to start on, to sort of experiment with what I could do with these languages. To try to find the happy medium between changing too much and changing not enough.

So we’ll see how that works out.

Let’s begin with the phonetics.

I’ve had the phonology of this language more or less set in stone for years; the real trick for them, however, is proving to be the orthography — that is, how to represent all those sounds without resorting to the IPA. I think I’ve found a solution for Hónskhardn, at least, but I’m not entirely satisfied with it still.

Hónskhardn has 29 consonants and 13 vowels. Not including allophones, length contrasts, and diphthongs. But we’ll still list most of those.

a /ɑ/ This is the “ah” at the back of the throat, as in English “father” or “pardon”.
á /ɑ:/ An accent mark indicates that a vowel is long. As in English “wad” or “sod”.
ae /æ/ This is the “ah” at the front of the mouth, as in English “bat” or “pass”. NEVER pronounce this as in “rate” or “Caesar”.
aé /æ:/ As in English “pad” or “bad”.
b /b/
d /d/
dh /ð/ Soft “th” sound, as in “this” and “those”.
dl /dɮ/ A voiced lateral affricate. Basically take “lh” from below and add it to a “d”.
dz /dz/
e /e/ This always has an “ey” sound, as in the word “freight”.
e /ɛ/ The “eh” in “get” or “wreck”. Yes, I know it’s weird to have <e> representing two different sounds. This is the main problem I’ve been facing.
é /e:/
f /ɸ/ Basically just the “f” sound, but more forward in the mouth.
g /g/ Always pronounced hard, like “get” or “gain”. NEVER as in “gin” or “gyro”.
gh /ɣ/ A non-English sound produced by starting to produce a “g” and then releasing a continuous stream of air from the throat. Or, if you prefer, simultaneously pronouncing “g” and “h”.
h /h/
i /ɪ/ The “ih” in “gin” or “bit”.
í /i:/ The “ee” sound in “machine”. NEVER as in “bite”.
j /j/ Pronounced as in German; that is, it’s always the consonantal “y” sound.
j /ɥ/ Pronounced by producing the consonantal “y” sound but with rounded lips. Unlike <e> and <o>, I’m not concerned about this one using the same letter as /j/, since I’ve only had it appear in a couple of words total.
k /k/
kh /x/ As in the German “ch”. Produced by pronoucing “k” with a continuous stream of air from the throat. I’d like to use <x> to represent this, but I feel readers will very easily misinterpret it as “ks” instead.
l /l/ Always pronounced as light (“light”, “list”), never as dark (“poll”, “hill”), even at the end of syllables. For best results, make sure your tongue is touching your top front teeth when pronouncing this.
lh /ɬ/ Pronounced more or less by producing the “l” and “h” simultaneously. The equivalent of the Welsh “ll” sound.
ll /ʟ/ Basicaly the dark “l” sound, but darker. Produce by pronouncing an “l” but with the tongue in the position for producing a “k” or “g”.
m /m/
n /n/
ng /ŋ/ The “ng” sound from “sing” or “rang”. Not that this sound may occur at the beginning of syllables as well as at the end.
o /o/ Always as in “code” or “road”.
o /ɔ/ As in English “for” or “core”. Another problematic letter being used for an extra sound.
ó /o:/
oe /ø/ Yet another non-English sound, most similar to the German “ö”. Produced by pronoucning “ey” and then simultaneously rounding the lips.
oé /ø:/
p /p/
r /r/ Always trilled, NEVER as in the English “r” or the French/German “r”.
s /s/
t /t/
th /θ/ The hard “th” sound, as in “thin” and “thimble.”
tl /tɬ/ Produced by pronouncing the “lh” sound after a “t”.
ts /ts/ Pronounced in full, as in German “z”; NEVER as English speakers pronounce “tsumani”.
u /ʊ/ The sound in “good” and “should”.
ú /u:/ The “oo” in “uber” or “blue”. NEVER with the “y” sound in front of it, as in “cute” or “huge”.
v /β/ See note on “f”.
w /w/
y /ʏ/ A non-English sound, pronounced like the German “ü”. Basically, pronounced as in “ih”, but with the lips rounded.
ý /y:/ Pronounced as in “ee” but with the lips rounded.
z /z/

Additionally, Hónskhardn has five diphthongs and four syllabic consonants.

Diphthongs: ai (“aye”), au (“cow”), ei (“hey”), oi (“toy”), and ou (“tow”).

Syllabics: m, n, ng, r. These only appear at the end of words, in more or less the same fashion as in English.

The stress is pretty simple: it always falls on the last syllable, unless that syllable is a syllabic consonant; then it falls on the second-to-last syllable.

So the name of the language is pronounced hon-SKHAR-dun. (hone-SCAR-dun, but with the “c” replaced with a velar fricative?)

My main concern is with “o” and “e”, since I feel the orthography should really distinguish between /o/ and /ɔ/, and between /e/ and /ɛ/. But I’m not sure how to do that without having to get into some really crazy digraphs or using more diacritics.

Thoughts? I guess?

Fabrication: Biocrystal Life Cycle

So I’ve had some time to reflect on my last post, and some responses to it.

One of my commentors mentioned that they liked the idea of parasites, and suggested that a compound naturally fatal to biocrystal might also be a good weakness.

Another commentor brought up the idea that, if biocrystal is supporting itself on local sand/stone and expelling a waste product back into the environment, some of that waste might end up in the biocrystal itself, building up and making the biocrystal less efficient, until the waste eventually just kills off the biocrystal.

Personally, I find I really like the latter idea, as it gives a definitive lifetime to biocrystal. It also allows for some interesting interactions.

That’s not to say there wouldn’t be parasites and other creatures that feed off of biocrystal — that’s definitely something that would be present, whether it be bacteria or viruses or integrated biocrystal beings that feed on biocrystal as an herbivore would on a plant. But these wouldn’t be the primary means of culling biocrystal.

So a slow atrophy is probably the best way for biocrystal to die. The speed at which this happens will probably depend on the size of the sample, with larger subjects dying off more slowly than smaller subjects (all else being equal).

Furthermore, this atrophy would begin in the core of a sample and spread outward. That way, new pieces budding off of a structure are unaffected and the surface from which the buds form would also be unaffected — at least until the end.

Dead biocrystal would then quickly fall to the elements, no longer able to sustain or repair itself. Any buds it had produced by the time of its death would break off then and start the cycle all over again. Either that or they would consume their dead progenitor in lieu of standard material.

But I also think that the matieral biocrystal consumes would affect the speed at which it atrophies and decays. Material rich in carbon (such as diamond) or silicon would be the most efficient, while materials with naturally radioactive elements would be the least efficient. That being said, it feels like a totally decayed piece of biocrystal should be bad for living biocrystal, or at least not optimally efficient, but it also feels like this would be an excellent way for the biocrystal ecosystem to recycle itself. But maybe it would be better for scavengers to consume dead pieces and turn them into something else?

To those who responded last time, thanks! Having someone to bounce ideas off of has been helpful. If you have further thoughts or want more clarification or such, please comment.

Fabrication: Animal Life

So sentient biocrystal.

I plan on there being two varieties: pure biocrystal sentients, and integrated biocrystal sentients. The latter of these two varieties would be far more common, if only because most biocrystal creatures wouldn’t need to move around much and thus would be better categorized as plants, and because the cerulite networks necessary to create sentience would be far more complicated than the biological neural networks necessary.

Pure biocrystal sentients (purists) would only really exist in environments where the nutrients necessary to persist and produce offspring were hard to come by. And seeing as biocrystal creatures feed off of sand and rocks, I can’t imagine many places where such creatures would exist. Perhaps in the depths of the ocean, or in particularly dense vegetation. Though, given biocrystal is supposed to also be able to somewhat feed off of organic matter, they might not even appear in some of these areas.

These purists would be composed primarily of large portions of rosete, and probably they would be generally snake-like in form, or perhaps even lizard-like.

Integrated biocrystal sentients (integrationists) would be far more common, and probably far more interesting. After all, technically these would be composite beings, a symbiotic relationship in a single form; the host would have to find some way to pass on the biocrystal symbiotes on to their children.

Probably all animal kingdoms would contain integrationists, though my primary conception of such creatures are generally reptilian — or more accurately, dinosaurish.

I imagine kinds of stegasaurs or dimetrodons with white albate sails along their spine, or large tortoises covered in black melanite, possibly pattered with albate as well.

I could probably come up with some other kinds of these creatures, but it’s getting late and I have to save *something* for my next post(s).

Questions? Want more details? Comment! Maybe I’ll come up with some answers or something cool in this system!

Fabrication: Manifestations 1

So after a hiatus of like a month or more, I should probably get around to considering devices and creatures that would integrate biocrystal, or else be made more or less entirely of biocrystal.

So I suppose I’ll be giving a few literary sketches of what these constructs would look like. I’ve already considered some such constructs in previous posts, and I’ll probably bring those in as we continue.

The first construct I’d like to go over is a literal power plant. I’ve primarily envisoned this as a tree, though it needn’t necessarily be in that shape, rather than, say, a bush. The central trunk would probably have an exterior of stone or incolorite — really, whatever the plant could get its metaphorical hands on — in order to provide protection from predators and the elements. Further up, where the trunk splits into branches, however, this covering would become less frequent or even nonexistant, the inner strands of ianthite blooming out in fractal threads. At the end of these strands, then, would be small melanite flowers, like a fruit tree in the early spring — only if all the flowers were black, rather than some bright/warm color.

I suppose it would be a rather strange sight, black flowers blossoming from violet branches bursting out of a translucent or stoney trunk.

Within the trunk, then, would lay the heart of this creature, a knot of cerulite at the intersection of all the ianthite radiating through the tree.

I initially imagined this tree being built by people for people, but I don’t think I’d be surprised to see it in the wild. Honestly, for that it would only need a few modifications.

In the wild, then, the cerulite core would be wrapped in a coccoon of viridite — not entirely, at least not usually. On regular intervals, probably about once a year or so, or when the tree’s roots bring up more material than usual, rosete tendrils would close the coccoon and another adjacent coccoon (or, more likely, three or more adjacent coccoons), activating the viridite.

Actually, now that I think about it, the regular intervals would probably be however long it took to begin this reproductive process after the completion of the last one.

But anyway, using sand or stone collected in these extra coccoons, probably transformed by some aurantite further below the core into incolorite, the tree would construct fetal copies of its cerulite core. Rosete tendrils would then carry these out of the trunk onto the branches where, probably in a strong breeze, the natal cores would fall away into the wide world.

Of course, using that template, you could probably get a huge variety of biocrystal plants, all varying in their cerulite programming to form different flower shapes, different flower numbers, different heights and widths and volumes, different numbers of branches, different trunk compositions, and different methods of spreading their seeds.

On the one hand I don’t think they’d compete with plants much, since they wouldn’t benefit so much from soil, since they’d receive their “nutrients” from sand and stone, where normal plants wouldn’t grow. It would certainly make deserts and mountains more interesting on worlds with biocrystal, transforming these voids into crystal forests, to complement the biological forests in more welcoming environments. On the other hand, I’m not sure how these biocrystal plants would flourish underwater, though at least the biggest issue for any potential beings of this nature would be the lack of light at lower depths — places that would otherwise be perfect for them, wide plains teeming with silt and sand and stone. At the bare minimum, any biocrystal growing at the lowest depths would have to exchange their standard photosynthetic charge for some other force, or perhaps rely solely on heat charge.

Thoughts? Queries? Ideas?

Hopefully next week I’ll be writing about animals made of biocrystal or integrating biocrystal.

Fabrication: History #11

For the final installation of the histories, I’ll be considering the last biocrystal, ianthite.

This is the biocrystal of communication, or rather of sending signals — whether they be visual, audial, or by other means (the exact means would be determined by cerulite attached).

The best use I can think of for this would be as a kind of wire between different biocrystals, allowing for circuitry and some programming for coordinating biocrystals. Though the first uses that people would find for it would be for signals, like flashlights or smoke signals — not exactly smoke signals, that is, but probably beacons of some kind that could be used in about the same way as smoke signals — or alarms.

Potentially, the medium of the signal could be modified and concentrated into a weapon, or for uses other than communication. Ianthite that emits light could be modified to shoot a laser; ianthite that emits sound could shoot rays; perhaps even ianthite could be made into guns or other projectile weapons.

Thoughts?

Fabrication: Pallidition & History #7

Well, I finally figured out how pallidition is supposed to work.

Pallidition is a magnetic magic, that pulls objects towards it. However, in order to do this, it requires an essence of the substance that it attracts.

Rubrite also has some magnetic qualities, but only on objects they touch, not at a distance, and naturally moves those objects. Pallidite does not move objects, but rather they remain attached to its surface.

Being a very simple system, it would have a number of different uses once people figured out proper essences. It could be used to hang things up, keep things in place, and such. Theoretically, one could also use it to attract body parts out of their owner, at least at short range (I don’t imagine pallidition working at a range of more than a few feet, if that, although perhaps a melanite battery could expand the range of it, perhaps). Though wiring the essence properly would probably be tricky and not nearly as efficient as just using a spear or a gun.

I might say that pallidite would be used in all the places a magnet would be used, but the applications would at least be significantly expanded, owing to the fact that pallidite could be modified to work on any number of substances, not just metals.