CAVE TALK
Game by snoopy, Sabrina Yen-Ko, Noé Antonio Chicas-Aranda, Jeffery Cai
Overview:
KE-VI TU-KE mo ke zapo ke — sorry, let me rephrase!
Cave Talk is a collaborative game for 3–4 players (ages 12+, 60–90 minutes) where your group makes up a language from scratch and then tests whether it can describe things nobody has talked about yet.
We wanted to explore how people learn a language and how that language evolves to fit the needs of those who use it. Since language evolution isn’t often explored in educational games, we wanted players to experience it firsthand by creating and adapting their own language.
In Cave Talk, you draw syllables at random from a bag and place them under cave paintings of colored rocks, walls, sticks, and cavemen performing different actions. A syllable could mean a color, an object, an action, or even an entire scene. It’s up to your group to figure out what each sound means and whether everyone agrees.
Once the board has some vocabulary on it, it’s time to put it to the test! Players take turns as the Speaker, drawing a secret goal nobody else has seen and getting two minutes to communicate it using only the syllables on the board and body language. Everyone else acts out what they think they heard, grabbing the right red, blue, or yellow rock, wall, or stick, or playing the role of a caveman. If they get it right, the goal goes in the Dictionary, and the team scores a point. If time runs out, the Speaker adds one more syllable to fill in whatever was missing.
After everyone has had a turn as the Speaker, the board is cleared, the syllables go back in the bag, and only the Dictionary carries over to the next round. Players must build on what they’ve learned while finding new ways to express ideas their vocabulary can’t yet describe.
Fellowship and challenge are the main kinds of fun we’re going for. We wanted to capture the energy of a party game, where players laugh, experiment with silly sounds, and physically act out their interpretations. At the same time, the growing vocabulary creates a puzzle that rewards creativity and collaboration. Through these interactions, we hope players strengthen their ability to interpret unfamiliar language, negotiate shared meaning, and adapt when communication breaks down, while gaining confidence in their ability to communicate without a common language.
Assessment Goals
Specific Learning Outcomes and Assessment Questions
Our initial learning objective was to increase self-efficacy when communicating in an unfamiliar language, and to develop the implicit skills necessary to facilitate this transfer of ideas (see Initial Ideation section below). As our game developed, we noticed that more specific linguistic ideas were being conveyed through the emergent dynamics of our game, and we wanted to develop our assessment to additionally capture these learning outcomes.
During gameplay, players first need to agree over shared meanings before they can move onto testing their language. This builds players’ self-efficacy when negotiating differences in interpretation and develops the implicit skill of being adaptable and flexible when interacting with others.
Additionally, since the number of usable syllables were purposefully limited, this created the necessity to use other non-verbal forms of communication to supplement the created spoken language. This developed players’ explicit skills of using context, gestures, and prosody (e.g. cadence, pitch, etc.) to supplement their communication. On a strategy level, this also developed the implicit skill of creating a complex system with heavy constraints. A potential side effect was the information gained by enacting out a scenario that illuminates why homophones form in language.
To target an outcome of increasing player information on language formation, gameplay was structured over multiple rounds in such a way that players had to constantly interact with language parts (choosing which parts of the language to keep, modify, or discard). We wanted to let players experience why language shifts over time, and under what circumstances new vocabulary is added to language.
We created eight open-ended questions to try to probe players’ understanding on each of these outcomes. The questions and their learning outcome are as follows:
| Question | Q# | Learning Outcome |
| How can people establish a common meaning for a word neither person has encountered before? (1-2 sentences) | 1 | Negotiating shared meaning |
| A stranger who shares no language with you holds up a cup of water and says a new word. How could you figure out whether the word means “water,” “cup,” or “drink”? Describe what you would do. (1-2 sentences)
And… How confident would you feel resolving the previous scenario? (1-5) |
2 | Negotiating differences in interpretation |
| How would you describe an unfamiliar scenario with only 4 words? (1-2 sentences) | 3 | Creating system with heavy constraints |
| Why might homophones exist in language? (1-2 sentences) | 4 | Information gain: homophones, evolution of language |
| How does prosodic information in your speech (e.g. pitch, stress, emphasis, loudness, tempo, phrasing) impact the meaning of language? (1-2 sentences) | 5 | Explicit skill: prosody |
| If two people interpret the same word differently, what could they do to resolve the misunderstanding? (1-2 sentences) | 6 | Negotiating differences in interpretation, adaptation and flexibility |
| What do you think is necessary for a group of people to develop a functioning language? (1-2 sentences) | 7 | Information gain: evolution of language |
| What necessitates the evolution of language? (1-2 sentences) | 8 | Information gain: evolution of language |
Summary of Key Findings
We collected assessment responses from nine individuals over our last three playtests. The Excel sheet of responses can be found here. Going question by question, we’ll provide a quick summary on the results by question:
| Q# | Learning Outcome | Trend |
| 1 | Negotiating shared meaning | Some players cited more systematic and concrete approaches to proposing, debating, and settling on a new idea. But many people’s answers stayed the same. |
| 2 | Negotiating differences in interpretation, self-efficacy | Average confidence increased from 3.44 to 4.22, indicating a moderate increase in self-efficacy. People mostly reinforced their ideas for the open-ended part. |
| 3 | Creating complex systems under heavy constraints | Generally players refined their understanding in a practical scenario. Pre-assessment answers were more broad, but post-assessment answers had specific and concrete solutions. Henri identified concretely that common words should get shorter representations. William identified key parts of language that need to be separated: colors, actions, and spatial relations. |
| 4 | Information gain: homophones | Players who were confused about homophones initially realized that it can be efficient to assign multiple meanings to the same sound when context can distinguish meanings. But many players also came into the game with a decent understanding already. |
| 5 | Explicit skill: using prosody | Players came in with a general understanding of how prosody conveys emphasis and implicit information. After the game they identified that pauses and emphasis can be useful to separate parts of a sentence, which is helpful during negotiations of meaning across language. |
| 6 | Negotiating differences in interpretation, adaptability and flexibility | Players initially talked about using a reference word to mediate. In the post-assessment, players cited solutions that involved more compromise. Notably, William cited a great anecdote: he initially opposed defining a word for person in the game, then relented. Later on, that word helped to differentiate between objects and action goals. |
| 7 | Information gain: evolution of language | This question was likely too vague to gauge any improvement in understanding. Players cited external exigence and shared goals as examples both before and after. |
| 8 | Information gain: language evolution | Similarly, this question was also too vague to gauge understanding. Players were able to cite the need for cooperation, as well as innovation, as reasons both before and after. |
Overall, it seems that the most effective outcome delivered by Cave Talk was improving players’ self-efficacy in dealing with an unfamiliar language. Specifically, their confidence in negotiating differences in meaning and reinforcing shared understanding improved more than any other outcome. It could also be argued that players gained a more intuitive informational understanding of the features of language (e.g. homophones and prosody) through a practical demonstration of those concepts in our game.
Iteration History
Initial Ideation, Design, and MDAO
Initial Intended Aesthetics and Outcomes
We started our design process with a core outcome in mind: teaching players how to communicate nonverbally or with an unfamiliar language. The idea arose out of how global our community is – Stanford students have, and will continue to, travel across the globe. When we visit unfamiliar places and cannot speak the language, the ability to communicate without using your mother tongue is indispensable.
In line with the taxonomy from Heather Browning’s Guidelines for Designing Effective Games as Clinical Interventions, our main goal was to design a game centered around the aesthetics of challenge and fellowship to put players in a scenario where they can practice communicating without English. By giving them exposure to the scenario, we hoped to increase the feelings of self-efficacy when players encounter unfamiliar languages and people in the future. On top of that, we wanted to strengthen players’ implicit skills in acquiring vocabulary in an unknown language . Ultimately, the game idea that generalized the best to all players was a game that would build up a language from scratch, where players could only communicate in that language.
Initial Mechanics and Intended Dynamics
We wanted a central mechanic to involve working with basic language components (e.g. phonemes, sounds) as they built their communal language. We hoped that this would lead to a more fluid and unconstrained dynamic when developing their languages.
The building-block mechanic for making the language comes from associating sounds to pictures. In addition to that, we needed to withhold a set of unseen pictures from the players to test their language. Both of these mechanics combine to form the dynamic of having concurrent development and evaluation of language.
Theming
The caveman theme naturally arose out of the concept of creating a language from scratch. The pictures could be either cave paintings or memories from the cavemen. The language would then naturally be one-syllable sounds that early humans could stumble into!
Color Choices and Accessibility
We made our three object colors red, blue and yellow because we believed they were the most color-blind accessible options, and we wanted to keep our game accessible to more people. We tried to take this one step further by giving each color its own texture (red is striped, blue is dashed, and yellow is plain), hoping this would make the game easier to play for people with different kinds of color blindness.
Playtest 1 – (10/4, external)
Players: Nanxi, Elana, Jay, Damian, William. All five players are undergraduate Stanford students, and avid board game players. These players are specifically our core target audience: they know how to maximize fun through both fellowship and challenge.
The first playtest confirmed that many of the intended initial dynamics and aesthetics were delivered. Players got comfortable enough throughout the game to have full-on arguments in their newly created language. However, the players’ tendency to min-max led to the dynamic where they anticipated all of the possible phrases they would need to communicate. As a result, they carefully and slowly developed their language so that it was expressive enough to cover all of the options.
Reducing Syllables
As a result, the training round was incredibly slow, but they immediately finished the testing round once their language was complete. To address this dynamic, we reduced the number of syllable tokens placed during the training phase in hopes of expediting the training phase.
We also wanted to force players to think on-the-fly and fill in some gaps in their language during testing. That would reinforce our language development outcome better, paralleling when languages have to quickly adapt to describe rapid innovations and changes.
Removing Text
A negative dynamic was when they started directly mapping syllables to words in a sentence structure identical to English. We attributed this to the fact that the cards had alt text that described the contents of the card, all in the identical structure “[verb] [X] with [Y]”. This inherently goes against our initial learning outcome of conveying how language naturally develops, because players are applying a pre-existing notion of language structure onto the game. To avoid this unwanted dynamic, we removed the textual description of scenes.
Removing Make Fire
The noun “fire” was only ever used in the construction “make fire”. This created a noun that basically masqueraded as a verb-noun combination. A campfire, as we drew it, is also more of a compound object, which was inconsistent with the other three nouns in the game. Not to mention, it wasn’t abundantly clear what performing the action of “making fire” would look like. Ultimately, we decided it was best to remove the idea of “make fire” from the game.
Removing Color Textures
We removed the stripes and dashes from colors after this playtest. We wanted to keep the objects and cards as streamlined as possible for most players, so the colors are now told apart by color alone. This trades some accessibility for simplicity, which was needed for our game.
Playtests 2 and 3 – (10/5, in class)
2nd Playtest Players: Yuina, Christina, Alexander, Xingyi. All players were fellow designers from CS 377G. They fall within our broader range of target audience given their age and general interest in games, but they may or may not fall into the specific group of people who love both fellowship and challenge games.
3rd Playtest Players: Olu, Lucas, Paprika. All three players are fellow designers in CS 377G – Olu and Paprika are students and Lucas is a course assistant. Lucas is definitely part of the core target audience as an enjoyer of both challenge and fellowship games. The other two are part of the broader target audience as students interested in board games.
Playtesting with a broader range of players showed us that the dynamic of trying to fully solve the language in one go is not universal. These players weren’t immediately aware of all of the possibilities, and were instead eager to move onto testing to see the strengths and weaknesses of their language.
Multiple Rounds
To support the observed dynamic without hurting our core target audience, we decided to add multiple rounds and make each round shorter. Early rounds would ease people into the game; later rounds would remain difficult to upkeep fun through challenge. We hoped players would open up and be more comfortable once they started testing their language, making for more fun through fellowship for less tightly-knit groups of friends.
Tutorial Round
A natural next step was to make the first round easier, so we made a tutorial round with colored objects only. Since players only had 6 unique ideas to communicate, we figured it would be an easy introduction to how the game works.
Even Fewer Syllables
We found that players in all of the playtests still consistently had too many syllables to work with, so we also lowered the number of syllables again to expedite gameplay. This way, the earliest round could act as the players’ introduction to the game and also to each other.
Playtest 4 – (10/6, external)
Players: Preston and Rachel. Both are undergrad Stanford CS students. Both of them enjoy games of fellowship and challenge, so they’re both in our core target audience.
In this playtest, we found that fewer syllables led to another emergent learning outcome. Players naturally ran into collisions in words because the number of syllables was fewer than the number of unique ideas. Although simplified, this gave players an intuition on why homophones may arise in language. These collisions also naturally encouraged negotiation over interpretations of language, as players individually developed opinions over how multi-syllabic words should be formed.
Faster Initial Training
Because players enjoyed the difficulty in expressing cards they hadn’t seen before, we wanted players to more quickly experience this difficulty by testing their language early on in the game.
Progressively Expanding Board
To do so, we experimented with a new progressive round system, where players initially started with six Scene Cards, rather than the previous nine. They would then play only one round of syllable tokens and then would begin translating Scene Cards. Upon failing a Scene Card, players would deal three more Scene Cards (up to a maximum of nine available on the board). Once successfully completing three Scene Cards, players would start a new round by clearing the board and dealing nine more Scene Cards. Players would then repeat until the deck concluded. The goal was to place players in scenarios where they were more likely to fail to encourage rapid iteration, as well as increase the challenge of the game.
Creative Gathering Tasks (implemented on Saturday 10/10 playtest)
This playtest gave us direction on how to implement the idea of having more difficult rounds later in the game. Rachel was, surprisingly, able to communicate the shape layout of a stickman with only the words for “rock”, “stick”, and body language. This showed us that ambitious challenge-driven players could communicate complex arrangements of objects.
To implement this, we created various complex arrangements of our objects, such as “Minecraft pickaxe recipe” and “bridge”. We thought these made for a challenging test after players figured out colored objects, but they were too complex to be useful in training. So, we limited them to the testing phase in Round 1.
Playtest 5: (10/7, in class)
Players: Olu, Yuina, Rachel, Aiden, Xingyi, and Emma. All players are fellow designers in 377G and fall in the broader target audience.
The experimental progressive round mechanic led to mixed outcomes:
- The new rule was more involved and required players to actively remember to change the board. Players had a hard time remembering how to change out the Scene Cards.
- Yet, players were still able to fill in gaps in their language despite having fewer Scene Cards. They used humming and body language to communicate an idea they had no word for.
Reverting Progressively Expanding Board, Reduced Scenes in Round 1
This playtest showed that progressive rounds added unnecessary complexity. It took up players’ working memory to process the rules, when we instead wanted them focused on strategy and internalizing their language.
However, players being able to communicate with fewer Scene Cards gave us more affordances when adjusting our rounds. Therefore, we changed the first round to have seven out of nine of the object cards. By holding out on two objects, players are confronted with two new ideas during testing, which serves our tutorial better.
The Dictionary, Success/Failure, Language Drift
Another idea we had to lessen the load on players’ working memory was adding a dictionary: a persistent written reference for the language. Whenever players succeeded in a test round, we would allow them to add the goal scene card into the dictionary and agree upon a phrase to transcribe next to it. Only adding successful goals to the dictionary would keep it as a correct representation of the history of their language.
To balance out the new idea, we also wanted to add a mechanic to help the language when players fail during testing. We reasoned that the most likely cause for failure would be an unexpressive language, so we let players place an additional syllable token if they failed during testing.
Players also tended to fully describe at least one scene on the board with their syllables. We wanted players’ efforts to not go to waste between rounds, so we added language drift, a phase where players add one Scene Card from the board to the dictionary at the end of the round.
Developing a dictionary also illustrates another nuanced learning objective: the evolution of language over time. It also adds another dimension when forcing players to negotiate differences in meaning, as they would have to untangle disagreements both present and past.
Written Assessment
We found administering the assessments with a large group ineffective – one playtester would answer the question and other playtesters would merely agree rather than come up with their own answer. Hence, we created a Google Form to collect assessment responses. We also added many more open ended questions that targeted more refined versions of our learning objectives.
Manufacturing and Polish, Rulebook Refinement
To improve ease of play, we wanted to make game pieces larger and more physical. After this playtest, we were able to 3D print and paint our props, as well as laser cut the tokens.
Players frequently asked for clarifications while learning the game. We needed to ensure playability out-of-the-box, so we began formalizing the rulebook.
Playtest 6 – (10/10, external)
Players: Siddarth, Henri, Diego. All three are Stanford students. They didn’t indicate a particular like for games of fellowship and challenge, so they belong in our broader (not core) target audience.
In this playtest, we saw that players used and referred to the dictionary throughout the game. We also saw that using new creative gathering tasks added a new layer of difficulty (see Playtest 4 iterations) by introducing positioning, an idea that could not be developed during training.
Game Board
When playing with the laser-cut tokens, players started placing them on the graphics within the Scene Cards and obstructing the scene. And since the token’s spatial position encoded a correspondence to an object, players used less body language, which detracted from our fellowship experience. To amend this, we made a game board that had a space to hold the Syllable Tokens beneath each Scene Card slot.
Easy/Intermediate/Difficult Split, Card Ideation
In this playtest, we saw that our original difficulty split was unbalanced because Round 2’s topics (communicating two colored objects) were solved once the players could do Round 1 (single colored object).
We decided that we wanted to introduce simple actions earlier, in Round 1, because communicating those ideas would hopefully elicit players to use their body language earlier on in the game. To keep the number of unique concepts small for the introductory round, we moved the idea of color into the intermediate deck. To try to avoid having redundancy (when nouns appear in two forms: by itself and in an action Scene Card), we cut the number of Scene Cards to 5 in Round 1.
We wanted the intermediate to introduce a similar number of new ideas, so the new concepts in this round were the 3 colors and 3 new verbs. But working off of the already learned concepts from Round 1 allowed us to build a large deck of combinations and enhance the replayability of our game.
Naturally, we included the complex crafting recipe cards into the difficult deck, along with many verbs that are modifications to existing verbs. For example, we added “take” as a contrast to the existing verb “give”. We figured the main challenge would come from trying to express modified ideas without any new vocabulary, so we made the difficult deck an optional challenge where you skip the training phase.
Playtest 7 – (10/12, in class)
Players: Olu, Aiden, Judy. The players are fellow designers from 377G, so they were part of the broader target audience.
We observed that players took a long time to read our dense formal rules, despite our mechanics being simple at its core. In addition to the heavy pre-assessment, it took us more than 15 minutes to start playing the game. Due to the time constraints, we did not observe significant differences during gameplay.
Yellow to Dark Gold
We asked for visual feedback on the final design of our cards, and the main comment was that the yellow was hard to see. To fix this, we matched the yellow more closely to the dark gold we already used for the rocks and sticks!
More Visual Rulebook
To improve readability of the rules, we reduced dense text, added graphics, and added more section headings. We hoped it would make the rulebook both less daunting and more illustrative because there would be more examples of correct gameplay.
Final Playtest – (10/12, external)
Players: Elana, Nanxi, William. Players are all Stanford students, and they thoroughly enjoy fellowship-challenge games. We revisited the playtesters that most embodied our target audience to test our suite of changes.
YouTube Link: https://www.youtube.com/watch?v=tJmSqzJHfQ0
Timestamps
Skip the section from 10:06-11:00 – Elana went to answer a message.
12:50: “Are the thin rocks different from the thick rocks… OH these are walls!” – They were reminded that there were walls from both the materials page and the Law of the Cave section, which have pictures mapping real props to cards/ideas in the game. Showcases how visual clarity in the rulebook clears up confusion.
15:40: Round 1 starts
18:31: Elana laughs at “tutu tu” to mean “hit with stick.” Because of the limited lack of syllables, players create many similar sounding words.
Coining a combination word for “person,” then explaining it to the rest of the table
21:24 Nanxi draws “po”, which they have established to mean “wall.” However, there are no more “walls” on the map, so she is unsure where to place it.
22:31 William brings up that “mo” both means rock and person (as the recipient of being hit) and explains to the table.
23:57: Nanxi comes up with “pomo” to mean “person being hit”
25:25: Elana shakes her head in confusion as William and Nanxi attempt to clarify the meaning of “po” (wall) vs. “mo” (rock) vs. “pomo” (person)
26:15: Nanxi demonstrates with props and Elana gets it!
27:07: “li pomo” = “to lift a person”: Nanxi uses her new vocabulary to describe a new unseen scenario that they could now convey.
Invention of “popo:” “hit within one person” vs. “hit another person”
29:07: Players get a “po” and try to figure out where to place it again. They realize they are using the same word for “hit with object” and “hit two objects together” by miming with the props and translating their own actions.
29:22: Nanxi comes up with “popo” to mean “hit two objects together” (“combine”)
29:25: Elana expresses confusion at the many repeating syllables, pointing at the many places on the board where “po” is used – she implies “popo” doesn’t make sense in this case.
29:54: Nanxi uses an existing agreed upon sentence to clarify by miming and speaking it (“tutu tu mo” – stick hit rock), then shows the new sentence (“tutu popo mo”) to show the contrast.
30:31: William takes this example and expands it by demonstrating “tu” means hitting someone else (or something someone else has), but “popo” means hitting two things together.
30:49: Elana asks a question, using props to demonstrate her guess of the sentence, and the other players confirm the correct translation.
33:14: Players feel the bag and realize there are still more tokens to be played. Elana sighs and laughs with frustration as finding the right place to play each token is difficult.
33:35: William draws “mo” (rock). They don’t have a great place to put it, so Nanxi puts one of the prop rocks on the board and humorously suggests placing it there.
34:29: Nanxi points to the bag: she’s deduced the last syllable is “li” so they should think about their next turn strategically.
“Pomo” as human or “pomo” as human recipient
37:30: Nanxi suggests using “pomo” as a general word for human, instead of just as the recipient of an action, demonstrating with “pomo li” meaning “human lifts something”
37:49: Elana agrees and translates another card: “pomo tutu tu pomo” (I hit you with a stick)
38:03: Elana and Nanxi gesture to show “pomo” means everyone. William doesn’t want the “pomo” to be used, since he doesn’t see a use with the current scene cards.
39:55: William finally gives in.
43:59: “These are the most logical instructions ever, I think we got it” Visual changes to the rulebook clear up misunderstandings in Phase 3.
Hit vs. Give
46:06: Elana reveals her card and it surprises the table: she’s drawn the “real” hit card, which is different from the “hit” card that the players thought they were translating. They realize that they have possibly not seen all actions. The limited number of cards in the first round allows for unseen difficulty.
46:43: “pomo tutu tu pomo, pomo tutu tu pomo?” William expresses worry that two different cards have the same phrase.
47:05:“litu” is established as “giving” while “tu” remains to mean “hit” in order to clarify the difference. We see similar words are built from existing words.
47:50: William just gets “wall” as his prompt, which the players easily guess. The difference in difficulty is humorous. Nanxi and Elana proceed to get difficult prompts.
51:10: William draws again and gets the very simple “rock” again as his prompt, prompting humorous outrage as his prompts are so much easier.
Coining the first color
1:00:00: Nanxi draws “li” and considers where to place it. Elana mimes “li” (to lift), using the meaning from the previous round. Nanxi starts to translate one of the new cards, using the previous language from round one, while demonstrating new vocabulary.
1:00:03: Nanxi initially states “li” to be the color “yellow,” but Elana reacts in shock because “li” already means lift.
1:00:12: The table agrees upon “lili” meaning yellow, since “li” already has a meaning, using the props to demonstrate.
1:01:21: William uses “mmm” to represent an unknown word.
1:02:15: The players draw “tu” and use it to determine the rest of the words for color: “tuli” = blue, “moli” = red, based on their existing convention of “lili” = yellow.
1:02:45: They expand this definition to include the people as well, demonstrating with the prop tokens.
1:05:04: “We actually don’t need anymore… Shoot this is going to get harder!” Players “solve” round 2, but placing the rest of the tokens adds to the challenge.
1:07:20: Elana is confused to see “mo” (rock) placed on wall (“po”), but then realizes it means “moli” for “red rock.” She places “po” to clarify the meaning of “mo.” Future token placements in this round are similar: the words are established, but recalling all of the words based on the board remains difficult.
1:08:00: William practices the language with props, Elana helps correct his confusions and mistakes.
Round 2: Caveman Act
1:10:40: Nanxi is incredulous… because she got an easy card. She says “lili po” to mean “yellow wall” and the players quickly get it.
1:11:57: Elana once again gets a really difficult card – the recurring imbalance is really funny.
1:13:36: “That was close!” (pointing at timer). The added timer adds a real sense of urgency and potential at losing the card, which players previously didn’t experience.
1:15:06: Elana discusses her frustrations with outside of game drama, Nanxi responds in CAVE TALK with frustration with “pomo” which Elana echos using body language. The language created in game is iconic enough for players to use it to represent their abstract scenarios as well.
1:16:38: “This dictionary is good!” By the end of Round 2, players have a complex vocabulary that the dictionary helps them remember.
1:21:57: Nanxi says “moli pomo” to refer to a person in real life walking by in a red shirt! The other players laugh in understanding.
1:22:25: William comes up with the word “po tu” to describe “poke” (consistent with “litu” for give and “tu”) for hit. The time limit and lack of training forces him to use existing patterns to coin new terminology.
Action vs. object clarification with “po mo”
1:24:42: Nanxi incredulous at drawing one of the difficult cards
1:25:41: William asks “pomo pomo?” to ask whether the action has a person on it. Nanxi shakes her head no and repeats “pomo” in a downward pitch. William then narrows down that the scene card must be a complex combination of objects.
1:26:26: “Oh my god we did it” – overcoming a huge and unprecedented hurdle satisfying
Adaptation to communicate direction
1:27:12: Elana draws a Minecraft pickaxe recipe and laughs incredulously too
1:27:55: Another shake head “mm mm pomo” clarification to mean that there is no stick figure
1:28:14 – 1:29:36: A straight minute of trial and error with the positioning. They realize that they have no verbal notion for direction because it was never in training. William tries to adapt by trying to assign “li” (long stretch) to mean to his right.
1:30:01: Players (kind of jokingly) use “li”, “li li”, “li li li”, and “li li li li” along with thumb gestures to indicate direction
1:30:46: William uses the “li” coordinate system to get other players to arrange sticks (tutu) in a certain configuration. Elana extrapolates that three sticks in that configuration could likely mean a triangle.
1:31:55: Nanxi suggests a new word for the noun “stick triangle” in the language: “tu tu tu” for each of the three sides.
More collisions in language as game goes on
1:32:27: Nanxi successfully uses the “li tu” (to give) but in a different order to communicate the reverse prompt (to take).
1:32:52: Players suggest (kind of jokingly) to invert “li tu” (to give) to “tu li” (to take). But then they realized those syllables was exactly the same as “blue” from earlier.
Print and Play
Box Design
WUWU
POPO LILILI