The Big Pedagogy

The fifth of January dawned cold, dark and threatening, and by the time I saw Wifey off to work, the rain began to come down gently. I scurried back into the house, confident I was ready for the serious downpour that was coming. The roofs were all clear, and the gutters all clean.

I spent the whole morning upstairs bringing order to the inchoate mass of last year's financial records, dreary work for a dreary day. Soon enough, the storm hit full stride, and though it was only of middling intensity compared to what these skies are capable of, it didn't let up until after one o'clock, at which time I went downstairs to fix lunch.

I looked out the kitchen door at the backyard, and was startled to see this:

I'd expected to see some flooding, because the turnout at the top of the driveway has always backed up in anything more than a light rain ever since the driveway was built. My surprise was at the extent of the flooding. This was easily twice as bad as I had ever seen it. I was afraid that if the skies really opened up, the water would back up into the garage.

Before I ran out and started filling sandbags, I checked Weather Underground, and saw only very light rain forecast for the rest of the storm. So I went on with lunch as previously scheduled.

As I ate, I thought back over the history of the turnout. It came into being in December 2004 with the completion of the driveway. This was just in time for the winter of 2004-05. As it happens, this was the wettest one on record here in Pasadena. Don’t take my word for it, you could look it up.

We weren't living here then, but I remember it well because at that time I spent most of my waking hours here working. I remember that our contractor's foreman—let’s call him “Rudy”—was very concerned that the turnout area would flood in a storm. He said it was because he'd noticed that patch of ground was especially slow-draining.

I took his word for it without much thought, because at the time I was very busy preparing the upstairs bathroom for the installation of the fixtures, which Rudy was pressuring me to get done, like, yesterday. Funny thing about that: for all the prodding he gave me, it took him a few months to get those fixtures in after I finished my work there.

Rudy cut a small drainage notch in the curbing at the bottom of the turnout, and even went so far as to dig a big trench leading from that notch over to a low spot of the yard as a temporary measure. He told me that eventually I would have to install a catch basin near the notch and tie it into the drainage system to take care of the problem long-term.

The turnout area did back up in heavy rain that winter, but so did every low spot in the yard, and the water would sink into the ground a few hours after the rain stopped. Of course, a lot of water did flow down Rudy’s trench, which gave the excess water a much greater area in which to sink into the ground, but I didn't give that much thought. 

January 10, 2005: Rudy's trench, part 1. . . 

. . .and part 2. 

We were in the house by the next winter, which I seem to recall was about average rain-wise, although my memories of that winter are pretty sketchy because I was sick in bed for most of it. The entire backyard might have been under water at some point that winter, for all I knew.

After that, we went into an extended dry period, capped by an epic four-year drought. During the spring and summer of 2006, I found the trench cutting across the yard to be a hazard to navigation, so I filled it in halfway; subsequent erosion and my leveling of the ground under the gazebo all but obliterated it.

The long succession of dry winters effectively removed the drainage problem as an active concern, and I found myself able to ignore it completely as I focused upon the exterior restoration work. The turnout still backed up in the rain, and I noticed that as the years progressed the backup became larger and slower to recede, but this was never more than a minor nuisance until now.

I did have to dig out the well around the persimmon several times through the years as erosion filled it up, and over time this erosion reached all the way to the curb, creating a new escape route for the turnout overflow directly into the tree well. This did not entirely replace the drainage provided by Rudy’s trench, but it certainly helped.

After I finished lunch, I continued to chew on these recollections as I finished putting our records into tidy piles and filing them away. I came to the conclusion that the backup was so much worse because it hadn't rained that hard for that long for at least a decade, and there was now no trench to carry the water away.

During the next 24 hours or so, it rained lightly and intermittently, and by the time the rain stopped for good, the tree well had drained completely, but the turnout was still quite flooded. The following 24 hours brought little improvement.

That completely baffled me. True, the drainage notch was clogged with dirt and debris, but even after I cleaned it out only a tiny trickle came through it, and at any rate the water should have sunk into the ground in that amount of time.

I dug a narrow trench from the notch over to the low spot, and blew the water down it with my big electric blower, leaving me with this:

It's not visible here, but there is still standing water in all the low spots. The turnout was so saturated that as soon as I blew the water out of a low spot, water would flow in from the surrounding area to fill it again. I decided to wait to investigate matters until it had dried out some.

In the meantime, I turned to the Internet to learn all I could about drainage. I knew something about the subject, but apparently not enough. I read about the importance of proper grading, and the various methods of moving water safely away from flood-prone areas.

I had a lot of time for study, because the turnout took eight days to dry enough to work with it, and it really didn't do much drying out until the sun came out at last on the seventh day and did its thing. Clearly, there was no downward movement of water in that area. It was as if there were a layer of concrete underneath it.

I did pick up a few clues during this time. I noticed that the wet area of the turnout was a darker color than the rest of the driveway. At first I thought that was due to its being wetter than the rest of the driveway, but the dogs were finding great fun in splashing through our little swamp, and I saw that at the bottom of the fresh divots made by their paws I could see the contrasting orangeish-tan color of decomposed granite. I recall now that these paw-divots had no water standing in them, which is what made them stand out from the surrounding muck.

At that point it finally dawned on me that the wet area of the turnout had a layer of dirt over it. That meant that water had all along been running into the turnout from the surrounding soil. Then I remembered that the backup had seemed to get larger and take longer to dry as time passed. That was enough evidence for me to conclude that the layer of dirt was what was causing the severe downward drainage problem in the turnout.

But how could the very dirt that allowed the passage of water right next to the turnout block it in the turnout? Here was a real mystery. After some heavy cogitating in the Lounge over a few pipefuls of black Cavendish, I concluded that the soil on the turnout must be mostly silt.

Loungers lounging in the Lounge.

Silt is the part of soil that is of the smallest particle size, and is thus the lightest; it stood to reason that it would move far more readily than the rest of the soil. Packed silt can be very resistant to water movement. I thus figured that the soil on the turnout was primarily silt.

I went out and looked at the grading of the soil around the turnout. There was a big rise, about six inches, at the bottom of the turnout that had been necessary at the time the driveway was built because there was a big liquidambar tree in that area that stood at that level.

December 2, 2004, completion day for the driveway.

Sadly, the tree died about five years ago of old age. We sure miss that tree. It was beautiful, and it gave us fall color all winter. In the South they call this kind of tree a sweet gum, and they grow like weeds, but here it was exotic and special.

Anyway, there is no further need for the rise.

The rise along the side of the turnout was not apparent to my eye. I had to put down some stakes and stretch a level line across the area and sight across that before the rise became evident. It crested at about two inches beyond the bottom of the turnout before it sank down into a low area that was the vestigial remains of Rudy’s old trench. The rise lessened toward the top of the turnout before it plunged down below level where the soil had eroded into the tree well.

With this information in hand, I had my plan of action: I would correct the soil grading around the turnout so that water flowed off of it instead of onto it, and remove the layer of silt. This would reduce the amount of water entering the turnout, restore the natural downward drainage, and prevent the re-formation of the silt layer. I hoped this would buy me the time I needed to finish the exterior restoration before taking any active measures such as installing that catch basin.

Next came triage. I decided to do the work in this order: remove the rise along the side, then remove the silt layer, then work on the large rise. I also resolved to build a proper well for the persimmon tree this coming summer, when I don’t have to worry about rain complicating the work.

I saved the large rise for last because it was the most complicated procedure, involving the most planning and work. I couldn’t simply remove all that dirt and establish a downward slope away from the turnout, because that would clash with the necessity of having the soil slope downward away from the house, and it would also require my removing the two nearby benches covered with potted plants and then having to re-level them when I put them back in place.

Oh, and I’d have to remove my new feeder pole, empty my new hole, dig it deeper to reflect the new grade level, and replace the pole. I didn’t expect to have nearly enough time before the next storm to get all that work completed, so I put that problem off until the end to give me time to come up with some sort of temporary fix that would prevent water from flowing onto the turnout from that area.

I put the removal of the silt layer from the turnout second because I had the romantic idea that if it were bone-dry, the silt layer would crack and come off like a hard candy shell. And so I started with the smaller rise.

Any soil work at the Farm House is likely to be difficult. It's not the kind of soil you can safely dig right into with a shovel or a spade, because it is hard clay shot full of rocks of varying size. You really have to loosen it up with a pick or mattock first, lest you dislocate your shoulder when your shovel hits a boulder.

The only effective way to establish a new grade in this soil is to dig up the area, screen it to remove all the rocks and the accumulated debris of 130 years, put the screened soil back down, rake and drag it out to the new grade, wet it, and roll it out with a heavy lawn roller. Then, over the following several weeks you have to keep raking it smooth, wetting it and letting it bake in the sun until it sets up.

A piece of buried treasure from the Farm House soil.

I wasn’t going to be doing all that anytime soon. I didn’t have the time, and I didn’t have the weather. It's a summertime job, because until the soil sets up any rain would oversaturate it and turn the path between the house and garage into a treacherous swamp. This would have been a worse problem than the one I was trying to correct.

Because of this, I had to try to loosen only the soil that I needed to remove, being careful to keep the remaining soil as undisturbed as possible. This would have been tricky enough without the many rocks embedded just below the surface. I tried a few different ways to get the job done, but I ended up having to chip the dirt away by chopping at it with a soil scraper. This is a tool with a triangular forged steel blade mounted on the end of a stick with a D-handle at the top.

I loosened the soil in a small area about a half-inch down in this way, which I then dragged into a pile with a hoe and scooped up with a spade. If I encountered a stone, I loosened it carefully with my big iron digger bar and removed it by hand. It was slow, tiring work, but it got easier as I developed the technique. The stones got larger and more plentiful as I moved towards the house, so I ended up having to disturb the soil there more than I liked.

When I was done, I had established a reasonably even slope going downwards from the house to the garage and just slightly downwards from the turnout to the low area across the yard. It wasn’t as much slope as the situation called for, but I wanted to see what happened in the next storm before I did more; there was a chance erosion would increase any slope I put in, and it was far easier to increase the slope later than to decrease it.

Next came the removal of the silt layer from the turnout. As I said, I expected this to be a quick job, but when I began to chip away at it with the scraper, I discovered that the silt layer was even harder than the soil. I took my soil tamper, an 8-inch square piece of heavy cast iron mounted on a stout wood pole, and pounded the silt layer with it in an attempt to loosen it and start to break it up. This had absolutely no effect.

It was really quite imaginative, my envisioning the silt layer’s coming off easily. It must have been the Italian in me. As I mentioned in my last post, "decomposed" granite (“DG”) is in fact crushed granite; it settles so firmly because its individual particles are quite jagged, and tend to interlock firmly when compacted. This same quality naturally caused it to become quite firmly interlocked with the much smaller particles of silt and clay in the dirt above it, once the dirt had dried.

I realized that to get the silt off, I would have to get it wet. I had wanted to avoid this, because when this stuff is wet it becomes quite slippery, and I was afraid that if I worked it when wet it would mix readily with the decomposed granite below it, thus making it necessary to remove a lot of the DG to get rid of all the silt.

That was when I remembered a procedure I had learned in the Doll House days, at our Culver City home, during the time when I first heard the ancestral call of the family green thumb. The soil there is black adobe, which is even harder than the soil here at the Farm House. The adobe, when it was completely dry, was extremely resistant to wetting; I would pour water on it, and the water would just bead up and sit on top as if the soil were tar.

I soon discovered that the only good way to get that soil to take water was to wet it evenly with water in which I had put a little dish detergent, sprinkled gently from a watering can. The detergent lowered the surface tension of the water, defeating its tendency to bead, and it would sink right into the adobe. After this treatment, the soil would then take water readily.

It occurred to me that I might use this same method to soften the silt enough to remove it without getting the DG wet. I filled up the watering can, mixed in a little Joy, and wet a six-foot square area evenly. As I had hoped, the water sunk right in. I waited a few minutes, then started scraping. An eighth of an inch of silt came off, and the rest was still completely dry. I filled the can with hot water this time, and put in a whole ounce of detergent. More or less the same result. Next time I used the same mixture but put down twice as much water, with only slightly better results.

I kept increasing the amount of detergent-spiked water and applying increasing amounts of elbow-grease, and I did manage to get down to the DG, working dry once I got down close to it. It took me several days to work my way stubbornly about a third of the way up the turnout until I finally admitted to myself that what I was doing obviously wasn't working well enough. I needed something that would break up the silt so the water would penetrate further.

I ditched the scraper and got out the dethatching rake, which has very thin, narrowly-spaced semi-circular tines. It's designed to pull thatch out of lawns, but its tines can be pivoted so that they act as tiny little soil-tillers. I didn't use it in leveling the rise because it's too aggressive, but that's just what I needed here.

It wasn't aggressive enough. It just slid back and forth on top of the silt. And with that, I tossed out the cautious, incremental approach. I could no longer afford it. By this time, another big storm had begun to loom in the forecast, and I had to start working more quickly. I realized that I'd have to get this silt good and muddy, and then just scrape it off the DG as carefully as I could. So I flooded the area with water from the hose and started raking vigorously.

After a bit of muddy drama, the water and silt suddenly resolved themselves into crumbly, dark, slightly moist soil, sitting on top of almost completely dry DG.

I had flooded that area, but some unseen force had kept all that water from moving out of the silt layer into the area directly below it—a force stronger than gravity.

Sounds like the Italian kicking up in me again, I know, but that was precisely what was happening. That, or some sort of ancient curse.

I swept all the soil up and got down on my knees for a closer look. Here's what I saw:

I said to myself, "IT BAFFLES SCIENCE!!" I could see that the moistened areas were where I had physically broken into the DG layer slightly; the bone-dry areas seemed to be right at the interface between the two areas. The interface between the two areas. . . .

A voice within me replied contemptuously, "No. No, it doesn't. It doesn't baffle science. It only baffles you."

It was another callback to the Doll House days, when I really had gotten in touch with my inner Oliver Wendell Douglas, when I lived for nurturing young growing things, when my workbench was usually covered with hundreds of tiny little pots full of little seedlings illuminated by grow lights on pulleys so I could adjust their height to account for growth. Back then, everything that I planted in our garden I had raised from seed.

Except for the roses. I tried, but I didn't have the skill to grow roses from seed. Roses I grew either bare-root or from pots. One thing that I learned early on about planting roses in that adobe was that you couldn't just dig a hole and plop the plant in.

Sure, you could do everything to ensure that the soil in the planting hole was perfectly blended, full of good things and well-drained; but once the roots had grown to the extent of that hole and hit that solid adobe, the roots would stop. Moreover, whatever water you put on the plant would stop at the extent of the hole. It was as if you'd planted the rose in a pot in the ground, a pot with no drainage holes.

When I first observed this situation, I consulted my gardening books and learned that this was called texture interface. As they explained it, when you put two types of soil with significantly different textures next to each other, a barrier to water passage occurs where the two soils meet.

I learned that to avoid this situation, you had to dig a hole much larger than the plant needed, and fill the excess with a 50/50 mixture of your prepared soil and the unamended soil from the hole in order to provide a more gradual transition between the two soils.

I was baffled by texture interface, and the gardening books really didn’t help much. They told me what it was and how to avoid it, but they gave no explanation of the science involved. I assumed it had something to do with fluid mechanics, but in what way I had no idea. I just figured that water was very skeptical of change.

What had been happening in the turnout was undeniably texture interface. As it turned out, the dirt on top of it was not all silt, but I'd say it was about 60% silt, and most of the rest was clay. So the dirt was extremely fine-textured, and DG is 100% sand, uniformly coarse-textured, and the result--well, the DG might as well have been concrete.

Here is a perfect illustration of the difference between rote learning and experiential learning, between learning by reading and learning by doing. Previously, I knew of texture interface because I had read about it and seen its effects indirectly, but I didn't really know what it was until I saw it happening with my own eyes.

Not that I'm knocking rote learning. It's how I’ve learned a lot of things like the times tables, the lyrics to the Roger Ramjet theme song, and lots of other useful, fun things like that. Rote learning is how the past prepares the present to face the future.

I was prepared for the hearty face-smack of epiphany by that bit of rote knowledge regarding texture interface that had been languishing in my brain, just waiting to be activated by experience. As Pasteur said, chance favors the prepared mind.

By chance, the very last big caper I had pulled was the installing of the bird feeder. Do you recall how I was able to re-use the sand I removed from the hole before I dug it deeper? I marveled at the time how cleanly that sand came out of the hole.

Down on my knees there in the turnout, bathing in the warm bright light of revelation, I looked over at the feeder, not more than a dozen feet away, and realized that texture interface had a big role in that job as well.

The sand came out of that hole so cleanly because texture interface works horizontally as well as vertically. It keeps the sand in the hole packed firmly together, holding the pole right in its place. Texture interface is in fact the reason why sand works nearly as effectively as concrete when setting poles in the ground. In our soil and climate, it arguably works more effectively than concrete, because it does not allow moisture to become trapped next to the pole, as concrete can.

It was but a small mental hop from that to understanding precisely why one puts a layer of small rocks at the bottom of the hole for the pole: to eliminate texture interface at the bottom. The rocks initially act as a buffer; the relatively large voids between them allow the water to fall out of the sand and then get absorbed by the soil. Soon, the soil and sand mix together in a gradual transition from one to the other to fill the large voids, thus eliminating texture interface.

This was all a great deal of fun, but it wasn’t enough. I was on a roll. I at last knew exactly what happens at a texture interface, but I still didn’t know how it happened. The mystery of how the silt layer could keep all that water captive until the sun finally set it free had not yet been solved. I got off my knees, went inside to my computer, and looked up a proper scientific explanation of texture interface, expecting it to be perplexingly counter-intuitive to match the actual physical event I had witnessed.

In fact, the explanation is perfectly intuitive and easy to understand. There are two forces involved, cohesion and adhesion, and while the names may be unfamiliar, the phenomena they represent are quite commonplace.

Cohesion is a force that attracts molecules of a liquid to each other. Water has high cohesion, which is what causes its relatively high surface tension.

Picture an icicle hanging off the eave of a house, just starting to melt in the sun. You know how a drop will start to form at the tip of the icicle, but it just hangs on until it finally gets too big, and then the drop falls off all at once, staying together until it hits the ground? That’s cohesion. The cohesion of water is strong enough to put up a good fight against gravity. Cohesion is why water beads up on the dry adobe soil of the Doll House.

Adhesion is the force that occurs when the molecules of two unlike materials have an attraction to each other. Liquids are strongly attracted to solid materials that contain very small voids. The combination of a liquid’s cohesion and its adhesion to such materials is called capillary action.

Picture what happens when you put a paper towel down on some spilled water on a countertop. The water goes upwards into the towel and spreads out until the water is completely absorbed. The water is attracted to the narrow voids between the many small fibers in the towel. Pick up the towel, and the water stays in it, held there by surface tension and capillary action, products of cohesion and adhesion, against the force of gravity.

This is just what was happening with the water in the silt layer. As with the paper towel, the voids between the tiny particles in the silt layer generated adhesion with the water that entered it, and it spread throughout the silt via capillary action until it was saturated.

Where the silt met the DG, at the texture interface, the water stopped moving because the voids in the DG are so much wider than those in the silt that capillary action stopped pulling the water forward at that point, and surface tension held it there. The water in the silt layer was held together by the combination of cohesion and adhesion so strongly that the force of gravity could not overcome it: the water in the silt did not weigh enough for gravity to pull it down.

It was as simple as that.

So I guess I had been right: water really is skeptical of change!

And with this, my spell of smartening-up was complete. It seemed as if I could feel my mind get suddenly less cluttered, as I threw out all the old, dusty, unconnected bits of soil lore that had been lying around in there. I didn’t need them any more, because now I know what texture interface is and why it happens. I’ve seen it with my own eyes, dispatched it with my own hands, made it do my bidding. I’ll recognize this perpetrator when I see it again in some new disguise. And I’m sure I will see it again.

Rote learning is a wonderful preparation for, and enhancement of, living. Experiential learning is living.

* * *

It was now a simple matter to remove the soil from the rest of the turnout, although I did have to remove some of the DG as well, to ensure I had removed all traces of any ancient curse, just in case. By that point I could see that there was a thin layer of soil on the driveway all the way across to the opposite curb and continuing down below the turnout for five feet or so, but I didn't have the time to take care of it, for the rain was coming in two days, and I had to do something about the big rise below the turnout before that.

Happily, by then I had a good plan for this: I dug a trench between the rise and the turnout. I tried to construct it so that water would flow into the trench from both the turnout and the rise, and the overflow from the trench would, if the rain was not coming down too fast, run out the side and down to the low spot on the other side of the yard.

With this, I put all my tools away and waited for the rain to come.

It came on schedule, and it didn't take too long before the turnout was backed up nearly as much as it had been the last time. I went out to watch how the water was flowing. Here's what it looked like:

Yep, I was right about having disturbed that area with all the rocks too much. As I feared, the water sank in there and oversaturated the area. And between this area and the tree well, I hadn’t established quite enough downward slope to move the water across the yard; my hoped-for erosion down to the low area did not occur. Instead, it mostly flowed into the tree well.

It was a mess, but I was happy with it. Sure, the water wasn’t going just where I wanted it to go, but it also wasn’t going where I didn’t want it to go. It wasn’t going from the soil onto the turnout, and that was the important part.

I can tell this is so by the fact that the water in the turnout is a lighter color than the water on the soil, as you can see clearly on the soil side of the drainage notch. Note the small finger of turnout-colored water flowing into the soil-colored water. This makes the distinction between the two colors clear, and confirms that no water is flowing the other way anywhere along the curb.

The swampy patch in the middle is a bit of an annoyance, but at least it is easily avoided, and in any event I can't really fix it until after the rainy season.

24 hours after the rain stopped, the turnout was free of standing water and was well on its way to being dry.

I was happy with this result, because it did allow me safely to postpone further measures until I had the time for them, but I was troubled by the apparent fact that despite my having accomplished essentially what I had set out to do, I had nonetheless failed to decrease the amount of water backing up in the turnout when it rained. This didn’t seem to make sense.

This time, it didn’t take more than about half a pipeful to come up with an answer: when I removed the silt from the turnout, I lowered the level of the turnout surface by over an inch. This caused the curb to act as a dam, holding the water back so that it had nowhere to go but down.

This restored my peace of mind for the other half-pipeful, but then a question popped up to disturb the peace: if I had stopped the flow of water from the soil into the turnout, then where did all that water in the turnout come from? It doesn’t rain more on the turnout than it does in the areas immediately surrounding it. Moreover, under the DG in the turnout, and in all the driveway, is four inches of coarse gravel. The turnout should thus have considerably better drainage than the surrounding soil.

True, Rudy did tell me that the soil under the turnout area was especially slow-draining, but I had finally dismissed this idea weeks earlier as highly implausible. What, was there a giant, turnout-sized boulder that just happened to be in that precise location, eight inches down? I figured the actual answer must be something else.

I looked back over the photos I had taken during the last rain and came upon a short video I took that I’d forgotten about. It showed me just where all that water in the turnout was coming from. I can show that to you here in another photo of the flooding in the last storm:

Look at the turnout below where the curb enters the photo from the right, and then follow along the curb to the left in the turnout. That is a flow of water from further back in the driveway. I didn’t record how far back this flow started, but the driveway continues to slope upwards from the turnout to the front of the house. This naturally creates a small but steady flow of water down that slope into the turnout.

This is why the turnout is a flood-prone area, and Rudy knew it. It was a direct result of errors in the way the contractor had built the driveway. Rudy lied to me because if he had told me the truth, we would have insisted the contractor install the drainage basin himself. It was his responsibility, because his work created the flooding problem. By planting the idea that the flooding problem was not the contractor’s fault in my head, keeping me swamped with work, and digging that big trench to mask how bad the problem really was, Rudy conned me into giving them a pass on it.

Yes, I was a chump. Don’t take my word for it. You could look it up.

Rudy was able to make me a chump because at that point I had not yet learned what a good liar he was. Over the remainder of our interaction with the contractor, I learned this all too well. During this time, it came to light that Rudy had been lying to me all along about crucial matters, and he had gotten the City inspectors and even our own independent inspector to go along with him. A bit of a sociopath, that Rudy.

It’s not as if I had not been paying close attention to the proceedings. I was here almost every day, working alongside Rudy and his men, asking questions, conducting my own inspections, bringing errors to Rudy’s attention. But I could not be here all the time, and I did not know everything about every aspect of the work.

I now know that whenever I was not here, and in every aspect of the work I did not know, Rudy was busy cutting corners and co-opting the very people whom I relied upon to keep him honest. When I caught him dead to rights, he managed to make it look like an honest error when he had no choice but to fix it, but otherwise he just flat-out lied.

It was a thorough and unforgettable lesson in the importance of maintaining a polite but unflagging skepticism in any business relationship. Rudy and water have taught me the power of skepticism.

Experiential learning. It’s the most effective way to learn, if you’re prepared for the lesson.

But it isn’t always pleasant.

* * *

A truly irresistible force.