Arguably, reading a novel requires a greater act of imagination than watching a movie — even the most thorough of textual descriptions won’t fully convey what a person or scene looks like. However, one could counter-argue a movie simply requires different acts of imagination.
A common movie technique is to make an “establishing shot” of a scene, then show a close-up — we are meant to imagine the people are still inhabiting the scene, even if we can’t see it. An internal monologue which might be fully expressed on a page might be merely implied by an actor’s facial expression. Watching the movie is an imaginative act, even if we’re unaware of it.
Adding graphics to computer games was a way of filling the “imagination gap”. In the process, though, other gaps were added, either inadvertently or by design.
The design of Mystery House wants the pictures to be our window in the world, and the text to be only incidental. Items that you can pick up are only conveyed by the picture.
Just at a glance, would you think from this scene that you can pick up a towel?
When look at this scene, do you think the middle of the room just contains some boards (as I did) or does it contain a sledgehammer?
Clearly, the idea here is equivalent to not writing out the actor’s internal monologue, but having them just act instead. Visuals mean aspects of the text can drop away. (Unfortunately in this case, it also means a lot of guessing what cryptic background objects might be called.)
…
The similarities to The Colonel’s Bequest continue: the initial cast of this game was: Tom, Sam, Sally, Dr. Green, Joe, Bill, and Daisy. So far I have found four of them dead, leaving Tom, Joe and Daisy. One of the bodies (Sally) had a blonde hair on it, suggesting the culprit was either Tom or Daisy. This reminds me of the one-clue-per-body plants that ran throughout most of Colonel’s Bequest (that game had the difference that the clue might have been either where the murder happened or where the body was later deposited).
Also, (again like Colonel’s Bequest) there are lots of ways to die which don’t seem to be related to the murderer. Turning on the stove in the kitchen results in it exploding. If you try to walk out of the dining room while holding a lit candle, you accidentally set the carpet on fire.
The most amusing death is reserved for trying to escape the house out the attic window:
OW
YOU FALL TO EARTH. LUCKILY YOU HAVE ONLY MINOR INJURIES. UNFORTUNATELY THE AMBULANCE DRIVER SMASHS INTO A VOLKWAGEN. NO SURVIVORS. YOU ARE DEAD.
In any case, I’m horribly stuck – the only thing I have resembling a puzzle I haven’t solved is a chest upstairs that needs a key I don’t have. I assume there’s some sort of secret passage activated by some graphical item in the background that I can’t decipher. I haven’t resorted to hints yet, but the lure of the walkthrough is strong with this one.

Especially when the parser is this frustrating. Argh.
Where are you stuck? It’s probably just a matter of just phrasing some command the right way or something in a picture you overlooked. That’s where most of the challenge in this game comes from.
I’ve managed some progress since I wrote this, and parser stupidity was involved. I will get into it when I am ready for my final post.
I’ve always found these very early Sierra games interesting from a technical perspective, if not very satisfying to actually play. Later illustrated text adventures would be just that: the pictures would serve only as non-essential eye candy. But here, Sierra tries to integrate the pictures with the game play, resulting in something that is at least as much a proto-graphic adventure as it is a text adventure. As you note, the pictures convey essential information, and they actually change over time, with objects you drop in a room being drawn into the room, etc. It doesn’t work all that great, largely because of the guess-the-word problems inherent in having a text-based input mechanism and a crudely drawn graphical output mechanism. Still, how interesting to note that this more ambitious approach was attempted before that of just sticking pictures into a traditional text adventure. As you probably know very well already, games in this style were everywhere on the Apple II for a few years before falling out of fashion about the time Sierra moved on to the King’s Quest style — which, when you realize how graphics-focused even these very early Sierra games were, suddenly doesn’t seem quite as much a departure.
If I was embarking on a project like yours, I would be hugely torn over whether these games rightly qualify as text adventures at all. Depends on what you prioritize in defining the genre, I suppose.
I think Demon’s Forge also had pictures-instead-of-words, but no other games like you describe are coming to me. (I was going to get into the difference more when I start the illustrated TRS-80 game which is the other candidate for First Graphical Adventure.)
I’m playing “All the Adventures”, not “All the Text Adventures”, so I haven’t even really thought about if it didn’t qualify. Interesting! (There are enough text-only bits I would say “yes”, like the whole sequence with the ambulance, but it does get borderline.)
The parser issues are interesting in that I have played some games with roughly equal fussiness in verb choice without as much issue. But since this game has hard-to-find verbs *and* nouns the combinatorial explosion makes the interaction much worse to deal with.
I totally agree about filling the imagination gap and how they ended up creating even more gaps in the process. Some of the images in this game are downright perplexing, leaving the player to wonder what on earth could have been going on in this house. Some of the time, the answer is just that the game designers didn’t know how to communicate something simple — for example, the fact that the victims were dressed up in the gear corresponding to their occupation is a hilariously awkward way of communicating who they are. If you wanted to interpret those images at face value, you might be inclined to think that the killer had some weird fetish with dressing up bodies.
I only just discovered your blog, by the way, but what I’ve read so far is very interesting. Nice work!
Thank you!
Mystery House is going to eventually make one more (brief) appearance when I play through the French translation.