Craig and I had hoped to give you a fully framed longhouse last week, but late into last Thursday night we had a massive panic attack. We had been spending the good part of two weeks debating, researching, consulting and re-researching possible roofing methodologies that would have been used when Iroquoian builders were constructing their houses. Yes, this seems to be also the main divergent of opinions as well between our key archaeological theorists; J.V. Wright, Mima Kapches and Dean Snow. As you may remember, Wright suggested that the longhouse building was separated into two structures; 80% supporting exterior walls and 20% separate rafters and roof. Kapches suggested that single continuous wall posts were bent over the superstructure frame with the ends terminating at the centre of the longhouse. Snow suggested that 60% of the longhouse was supporting wall and 40% a more arbour like but fully attached roofing system.
We have been attempting a hybrid system between Snow’s interpretation and Kapche’s version. Essentially a longhouse that has a 60/40 proportional split between wall and roof (Snow) and a continuous wall post that begins bending at about 60% of the height. What nobody has really not talked about is how the smoke holes where constructed or even if they were above the fire hearths inside! Thus our ability to understand how smoke holes were constructed or even if they were purposefully built into the roof during the construction process has been hindered by a lack of oral, historical or archaeological evidence.
One of the first problems is that the archaeological record clearly indicates that fire or cooking hearths can be found throughout the inside floorplan of a longhouse. Generally as Dodd indicated in her 1984 research, they tend to be grouped along the centre/middle of the longhouse (see image below). 
However, Varley and Cannon’s 1994 research on hearth spacing also indicates that hearths did move and there could have been both a cooking hearth and a heating hearth in close proximity to each other. Further, there is no indication that the number of hearth’s actually represented the number of family units within the longhouse (normal convention is that it is one hearth shared between two family units on either side of the longhouse bunking system). The excavation map below of the Lawson Site, which is informing our research, shows two fire hearths in House #5 somewhat inline along the centre of the house. However House #6 has one larger fire hearth one one end of the house and three smaller ones group on another. I want to urge caution as well. Just because they didn’t find evidence of additional fire hearths, it doesn’t mean there wasn’t any (they could have been scrapped away during the excavation or they could have been removed by the original occupants). It does help to visualize the problem of where to put the smoke holes if we were to reconstruct directly from the archaeological data.
In Ron Williamson, David Smith, Rodolphe Fecteau and Robert Pearce’s 1979 Ontario Archaeological Symposium paper entitled The Longhouse Experiment: An Experience In Iroquoian Archaeology, the four archaeologists spent 30hrs in a reconstructed longhouse at Ska-Nah-Dot in the middle of a January snowstorm. In -15C weather, the experimental archaeologists stayed in a 21.3m long, 6m wide and 4.5m high, wood framed and elm bark shingled longhouse. The house had four smoke vents along the centre of the roofline that were .4m’s in diameter. The smoke holes had hinged covers that could be opened with a pole from inside (Williamson et al., 1979). The interesting part was there were 5 fire/cooking hearths on the ground, measuring about .6m’s in diameter, running down the middle of the longhouse. We will return to this experiment over the next few weeks when we start recreating 3D smoke within the virtual environment, but for now, this description is all we have on the dimensions or physical make up of a smoke hole would be.
Further, in discussions with Neal Ferris, there is the question of weather the smoke hole would even indicate where the actual fire hearth would be below. Convention suggests that the smoke hole in the roof would be directly above, however Ferris has raised the point of rain or melting snow providing a constant dripping on the fire hearth below. If this was to happen, the fires would go out, but not before substantial smoke could accumulate. A cover, as briefly described above, would not be totally effective in keeping dripping water out, so my assumption would be that the smoke holes would be offset from the fire hearths below.
Lastly, Williamson et al’s description indicates a “diameter” of .4m’s. Longhouse construction is clearly linear and with limited fine cutting tools, I can’t see Iroquoian builders up on the top of a 7m high structure trying to cut a circular hole in the roof. I’m going to take an educated guess that the “smoke holes” were rectangular or squarish and that the builders would use a similar shaped piece of bark shingling to act as a cover.
Hence, Craig and I started reconstructing the roof line to be much flatter, allowing for square or rectangular smoke holes to be built into the the roof and rafter system. We drew some inspiration from the modern architectural drawings Ron Williamson provided in Longhouse 2.5 which provided a plausible method of smoke hole construction. The image below includes a second horizontal set of beams down the middle of the roof supports to act as a connection point for framing and strapping of smaller gauge poles.
You’ll notice we have also included the interior door frame construction. Again, from historical accounts there are indications that the flat inner doorway was constructed with the same lightweight cedar shingling that was mentioned in Snow’s account of bunk compartment walls. After some discussion with Craig on how they would hang the cedar strips, we concluded that a door frame had to be constructed to act as a brace. In the shot about Craig extended a horizontal pole at the bottom, but the door posts are assumed to have been dug similar to the other posts in the house. The archaeological record shows no trenching or vertical lines as soil stains on either end of most longhouses, so I told Craig to remove the poles touching the ground and just extend the support poles on the first set of benches.
In the downward shot above, it’s clear we would need smaller more lightweight supports to frame the smoke holes. We decided to use the same diameter poles that are being used for the exterior wall strapping (around 3cm’s) to help frame up the roof supports.
In the shot above, we start laying out the poles for the smoke hole framing and to act as supports for the top roof shingles when we start adding them to the structure.
It’s at this point when panic sets in. We did a frontal render of the longhouse and it seems somehow it has grown squatter and wider. In particular I was having issues with the exterior wall posts and the fact they just looked too elongated. Going back to previous renderings reveal that yes, the longhouse has been getting wider in width and shorted in height. 
As we know from oral and historical sources, longhouse height was the same measurement as longhouse width. Currently we are using Dodd’s research which suggests the average width for a Northern Iroquoian longhouse was 7m’s, thus the height would be 7m’s. Craig did a simple measurement rendering and clearly our assumptions proved correct! Somehow during the modelling process we had changed the dimensions without knowing.
Again, madly pouring through our data from Dodd, Wright, Snow and Kapches, we settled on resetting the longhouse so the width equals the height. Further we knew that the bunks were exactly 2m’s (or just below) in width on either side of the longhouse and that according to Dodd the centre width between the two bunking systems were typically 3m’s wide, which gave us 7m’s in total width (2m + 3m + 2m) and thus an acceptable range. Height again is purely speculative, but we’ve wanted to maintain the dimensions as mentioned in the oral and historical accounts.
A quick render proved the changes visually made sense width wise, however the height looked just too high. The door posts also looked way too thick, so we scaled down the diameter to be more inline with a slightly thicker version of the wall posts (see below). I also asked if we could lower the height by .5m’s to 6.4m in height instead of 6.9m’s (same as width).
The Image above shows all of our revisions in place. Stylistically (artistically) we lowered the height just slightly, changed the diameters of the door posts (which we are assuming was an opening of about 3ft x 6ft or shorter) and setup the roofing system to accommodate smoke holed along the centre line of the roof. At this point we have 8 family compartments, which based on normal convention should have one fire hearth for each pair. So a minimal total of 4 fire hearths and subsequent smoke holes. When refining the smoke holes next, they will be off-set from being directly under the centre point of where we think they would have normally put the fire hearths.
Now that we fixed one problem, we wanted to do a rough check to see if length was going to be within Dodd’s numbers. Each bunk is 4m’s in length and we have 4 along the length of the longhouse, so a total of 16m’s. We will be adding the vestibules (rounded half cigar shaped ends) to either end, which according to Dodd was another 4m’s in length, so a total 24m’s long (4m + 16m + 4m).
This round of changes was really an eye opener in terms of keeping track of dimensions within the 3D space. We were lucky to have discovered the dimension differences before getting further along the modelling process as it would have been more difficult to fix later. It made me think of the decisions the original longhouse builders would make and how they might fix or repair their structures during the building process and afterwards. Further, I’m being influenced by my artistic training I’m concerned that my visual perception might distort the archaeological data, so this is something that will need to be follow closely as we start creating the atmosphere and interactive pieces.
Disaster averted but a very beneficial process!
One Reply to “Longhouse 3.3”