Trouble Shooting a Garden

JL has a very large second-home garden – 500M2, in France. It was unwanted; but went with the house, which was going for a song. Irresistible or what? However, with 3 visits a year making a stay 2 of months in all, garden maintenance is a bit of a problem. It takes JL 2-4 hours day, when in residence and he is no spring chicken. The challenge is to maintain the garden as desired with an acceptable level of effort.

Meeting the challenge has not been easy, in-spite of JL being supported by the ideas of an ex-colleague and friend – AL, who happens to be a landscape gardener. The idea behind the post is two-fold. First, to open up the challenge to others, friends and readers; who are also interested in gardening. Second, to report attempts to add meaning to the designing and implementing of such a garden.

The process is as follows. JL posts photos, showing the state of the garden. Also, identified will be maintenance problems, their associated solutions and their effectiveness. Others are welcome to suggest ideas, intended to help meet the challenge.

Here is the original plan of the garden made by Doris Long – the gardener of the family.

Key F=Feature

The summer ’20 visit to Auzon followed a previous visit in September of last year, the Easter visit being cancelled.The garden, then, had not been maintained for 10 months or so.


Before – 2 Flower Beds (F8) and 2 Mosaics (F9) (Cochon d’Auzon; Chateau d’Auzon)

Before – 2 Flower Beds (F11) and 2 Mosaics (F12) (Coq Francais; British Lion))

Completely overgrown.

Earlier Problems – Excessive peony growth, weeds.

Earlier Solutions – Reduce peony groups and number per group, treat weeds.

Current Problems – Excessive peony growth. Weeds in mosaics.

Current Solutions – Flower beds – more tiles, more bark, reduce peony number per group. Spot treat weeds.

After – 2 Flower Beds and 2 Mosaics (F8 and F9) (Cochon and Chateau)

Any other ideas anyone? We must be told!

Exchange between JL and AL

New Concept of the ‘Weedy Garden Cottage’

  • ‘Weedy’, because their presence is now accepted as inevitable, given the time available to maintain the garden and the growth rate of the weeds.
  • ‘Garden Cottage’ in contrast to ‘Cottage Garden’. The latter is a type of garden. The former a type of cottage.
  •  NEW STRATEGY
  • All flower beds, mosaics and minor paths between:
  1. Problem=weeds
  2. Solution=plastic, gravel, bark, spot spray (Solution implies acceptability)
  3. A – So no intervention necessary?
  4. J – For as long as this works……..
  • All main paths 
  1. Problem=weeds
  2. Solution=plastic, gravel, clear, spray.

A – Not sure if TQ (Task Quality – how well the task is performed) and adequate and UC (User Costs – time and effort of the gardener – JL) acceptable in this case? If so, no intervention necessary?

J – Yes, OK so far.

  • Peach trees x 3.
  • Problem=excessive growth, UC unacceptable for cloud pruning.
  • Solution=cut back bifurcating trunks (not sure about this).

A – Probably better to do this during your autumn/winter visit, as summer pruning presents a greater risk of disease.
Perhaps also desirable  to discuss the method with reference to a picture of the tree(s), as approach may depend
on size of the branch for removal and how high up the bifurcation occurs. Iphone image should be fine.

J – leave as is for this at present.

  • Holly tree/bush/laburnum/wisteria (Doris and pergola)
  • Problem=growth.
  • Solution=rigorous pruning.
  • Terrain de Boules (F13)

A – Again, not sure if TQ adequate and UC acceptable in this case? If so, no intervention necessary?

J – Yes. So far, so good.

Terrain de Boules

       1. Problem=Weeds. Too many for me to clear on a single visit. Due to faster growth and JL slower workrate.

       2. Solution=Cover as much of the ground in whatever way possible.  Let weeds grow where they are acceptable, for example, in parts of the artefacts (or in the case of the boxes through them).

Herb and Flower Plot (F3)

         1. Problem=Weeds.

          2. Solution= add vegetables to the plot.

Contribution from Karen Davis (to see photos of her and Denis’ garden go to Garden Arts)

K. How often do you and Paola visit Auzon?  

J. About 3 months in all. 3 weeks at Easter. 7 weeks in the summer. 2 weeks in the Autumn.

K. How do the peach, laurel, fig trees, wisteria and iris get enough water to stay alive when you’re not there?

J. When it rains…… Seems to work. None has died. In fact, they all thrive. The soil is very fertile.

 K. In my opinion, accepting the weeds and letting them rule the garden is to embrace chaos.

J. Of course, I agree, but it depends how you look at it. Also, what you think a garden is and what it is for. What you think gardening is for you. How you may, or may be not, be willing to change your thinking and feeling with respect to garden and gardening.

 K. Are there any young people in Auzon (younger than 70 or 80), you could PAY once or twice a year to pull out the offending weeds – keeping the pretty ones if growing in acceptable location.

J. There are. I tried this, when Doris’ health was failing. Neither followed my instructions. They just do in the garden, what THEY think the garden should be and look like. Virginia grew flowers. Jacques just cut back the weeds the other side of the garden wall (not part of my garden).

K. Some of those weeds are very invasive and will ruin the peace of your garden by becoming a tangled mess.

J. Correct. I have photos to prove it.  

See above.

One or two people could probably get the weeds out in 1 – 2 hours?  Then you would have a tidy view of the garden allowing for clear analysis of what else is needed – likely some pruning and that’s it.  Garden now peaceful and zen-like letting your artistry with found items shin

J. Maybe. But see above. 2 people would not be able to agree what they should do, which in neither case would be what I want. Also, I enjoy gardening. Further, I am old fashioned, if you have a house – look after it. Same for the garden. Zen-like has given me an idea.

Will return to this.

Karen, thanks for this. I found it very interesting and helpful. A good test of my ideas and feeling about the garden.

The Current Scheme is not Working Well

Here is J’s attempt to rethink it a bit.

  1. The Auzon garden is a space.

       2. The space is made up of (garden) features. For example, flower beds, paths, terrain de boules, etc.

       3. The features are made up of states. For example, flowers, gravel, patio bricks, weeds, herbs, paving stones etc.

The criteria for the garden spaces remain Tq and Uc. 

The criteria for the JL experience remain Tq and Uc.

As an example:

Terrain de Boules

  • Garden space
  • Garden features – 3-D artefacts, such as metal hoops (filled with stones), wild flowers, weeds, compressed yellow sand etc.
  • Performance
  • Garden Tq good; Uc poor. Need to make eliminating large weeds easier.
  • JL experience Tq poor, Uc poor. Need to find weeds more attractive. Need to make experiencing weeds more pleasurable.
  1. The Auzon garden is a space.

       2. The space is made up of (garden) features. For example, flower beds, paths, terrain de boules, etc.

       3. The features are made up of states. For example, flowers, gravel, patio bricks, weeds, herbs, paving stones etc.

The criteria for the garden spaces remain Tq and Uc. 

The criteria for the JL experience remain Tq and Uc.

As an example:

Terrain de Boules

  • Garden space
  • Garden features – 3-D artefacts, such as metal hoops (filled with stones), wild flowers, weeds, compressed yellow sand etc.
  • Performance
  • Garden Tq good; Uc poor. Need to make eliminating large weeds easier.
  • JL experience Tq poor, Uc poor. Need to find weeds more attractive. Need to make experiencing weeds more pleasurable.

Exchange Between JL and AL on the Above

1. The Auzon garden is a space.

2. The space is made up of (garden) features. For example, flower beds, paths, terrain de boules, etc.

…so features are here distinguished at the level of sub-spaces in the garden…

J. Yes. For example, the TdeB contains artefacts. Also desirable weeds, but not undesirable ones.

3. The features are made up of states. For example, flowers, gravel, patio bricks, weeds, herbs, paving stones etc.

I’m not clear on how these states relate to the gardener’s work. Previously, I would have viewed entities such as gravel, patio bricks and paving stones as objects, for example, constituting a path, which might assume the state of ‘weedy’ or ‘weed-free’. The task of weeding the paths would transform the objects comprising the paths (gravel, paving stones…) from the weedy to the weed-free state, with a performance reflected in Tq and Uc. Here gravel and paving stones seem to be states in themselves… Am I missing something?

J. Your use of ‘objects’ could be retained and would be tidier relative to the D+J conception. Features then come in as the next level down. This change, however, does not change or resolve the problem. Acceptable and unacceptable weeds would still (and could be) distinguished.

The criteria for the garden spaces remain Tq and Uc. 

Are you referring here to the performance of the task of maintaining the garden spaces, or something more general?

‘Performance’ in any and all cases. No change there. See also the introduction earlier.

The criteria for the JL experience remain Tq and Uc.

Does the ‘JL experience’ refer to the experience of performing the (previous) garden space maintenance task? Perhaps including affective aspects….? Is this what distinguishes it from the preceding criteria?

J. Both could be included. The critical difference is that JL’s weed experience needs to be acuired and encouraged……JL can no longer experience wild violets as ‘just weeds’ to be weeded….. 

As an example:

Terrain de Boules

  • Garden space
  • Garden features – 3-D artefacts, such as metal hoops (filled with stones), wild flowers, weeds, compressed yellow sand etc.
  • Performance
  • Garden Tq good; Uc poor. Need to make eliminating large weeds easier.

Why is Garden Tq ‘good’ here? Is it something like ‘closeness to the specified state’, where the specification allows (some) weeds to grow?

J. I was not clear here. Tq is good in the absence of undesired weeds, that is, with only acceptable weeds.

  • JL experience Tq poor, Uc poor. Need to find weeds more attractive. Need to make experiencing weeds more pleasurable.

Why is JL’s experience Tq ‘poor’ here? Is this because the specification is not concordant with JL’s underlying/historical preference for a weed-free space?

J. I don’t yet experience acceptable weeds as acceptable. My first reaction is to eliminate all weeds. CognitivelI know this as an intent, but not otherwise.

I could do with a sketch of the elements of the new conceptualisation relative to the old one, I think…

J. I agree this would be good. The first move, however, is to agree on which end is up first……

At first sight, the problem seems to be all about the apparently dissonant representations of weeds that G  holds.

A.

  • The concept of the ‘weed’ has an intrinsically negative connotation, which complicates matters when discussing the possibility that they may have positive attributes. However, I think a general principle holds, that objects are able to carry both positive and negative attributes simultaneously.
  • Any plant, whether designated a ‘weed’ or not, will change in its appearance over time, i.e. it will change its visual attribute states. Some of these changes will be positive – like the appearance of flowers in the spring; while others might be perceived as negative – like growing too big for its space during the summer and overwhelming adjacent plants, or going brown in autumn. Thus, ‘desirability’ of an attribute state is often (always?) context-dependent and is based upon a value judgement. Any weed is likely to be more attractive/acceptable at some times in the year than at others.
  • In the D&L conception a task seeks to change the (attribute) state of a domain object from a less desired state to a more desired state. As a performance metric, the closer the state change achieved by the IWS is to that desired, the higher the Tq achieved. However, in terms of the conception, the ‘desirability’ of the state being sought in the conduct of a task is arbitrary  – it could be ‘good’ or’ bad’, dependent on the point of view of the beholder.
  • The human element (G in this case) undertaking a gardening task must perform an evaluative function, dynamically assessing progress towards the desired set of attribute states and adjusting their behaviour accordingly. This same evaluative function must presumably take place before work begins, in specifying the task in terms of the desired end state and an acceptable level of costs to G.

One might speculate that the mental representation of an object held by an individual may embody positive and negative connotations, established and potentially changing, over time. However, some of these connotations may be stronger than others as a consequence of education and/or particularly salient learning experiences. These stronger connotations might manifest in the way that the object is habitually perceived by that individual.

One might further speculate that G will have established a representation of the object weed over his multi-decade experience of gardening, during which weeds have been repeatedly reinforced as having attributes with negative connotations, such as rapid growth and spread;  often deep rooting and so hard to extractintrusive/invasive habits; as a group, usually visually unattractive in the garden; disruptive of intended plans for the space. In the context of these negative attributes, weeds detract from G’s positive experience of the garden and hence reduce Tq. Furthermore, their necessary removal, again given their negative attributes of proliferation and tenacity, are associated with high Uc.

In this context, it would be unsurprising that a changed approach to the management of the garden, involving a more nuanced connotation of weeds and their management, might give rise to conflicting experiences of this class of plants for G. Specifically, G’s conception of weeds as ‘bad’ has been established and reinforced over decades. Furthermore, G’s natural inclination to favour order and tidiness over (even managed) ‘disorder’ adds to the reinforcement of the view of weeds as ‘bad’.

Implications:

  • Choice of vocabulary in specifying a task for analysis in IWS terms needs to be neutral, to avoid  inherent biases in connotation. Alternatively – unless it is self-evident – the particular connotation attached to an attribute whose state is to be changed needs to be made explicit.

A weed might be more neutrally defined as a  self-propagated native plant (SPNP?!) or perhaps just a ‘wild plant’. These could be specifically classified as being good, bad or conditionally welcome in a specific context.

Note that there are several beneficial consequences of allowing an increased proportion of wild plants in the garden, such as more biodiversity and less detrimental impact on wildlife in the vicinity resulting from pesticide use, in addition to reduced physical Uc.

  • Discussion point: Can performance be specified without some understanding of the mental representations held by the human component of an IWS?

On the basis of G’s experience, it looks as if there might be a requirement to make explicit a model of the domain as held by the human element of the IWS. Perhaps this could be quite crude yet could still assist in defining more explicitly/precisely the change in the state of the domain that would reflect high and low Tq for a given task.

G might hold a (probably small) number of  internal representations [templates?] of the garden reflecting alternative potential [desired] formations of the space. These might be historical – how the garden was originally envisaged; how it was when JL assumed sole responsibility for maintaining it. Some might be more recent – the garden under a well implemented low maintenance regime.  Representations of plans for the garden at some future time would also be possible – for example  the TdeB accommodating a swimming pool or the garden…

The set of garden features might be the same in most of the representations, but representations of alternative formations would differ with respect to the constituent objects  and their desired attribute states. The default representation utilised when doing actual work in the garden would be that of the current desired formation of the garden.

The issue raised by G in relation to weeds might reflect interference between a historical representation and the current working (default) representation of the garden domain.

  • Discussion point: How might the mental model of the human element in an IWS be represented and operationalised?

I have some thoughts, but probably premature to consider at this point.

  • Discussion point: How to deal with G’s discomfort? Maybe changing the label ‘weed’ to ‘self-propagated native plant’ wouldn’t be sufficient….. Perhaps reframing garden tasks in terms of new ultimate goals for the garden, such as improved environmental sustainability and increased natural diversity would be more fruitful?

Auzon.

Les hameaux

JL and AL Continue


At first sight, the problem seems to be all about the apparently dissonant representations of weeds that G  holds.

JL1 Agreed. Dissonance is not helpful for the matters in hand. Needs sorting out.

Observations:

  • The concept of the ‘weed’ has an intrinsically negative connotation, which complicates matters when discussing the possibility that they may have positive attributes. However, I think a general principle holds, that objects are able to carry both positive and negative attributes simultaneously.

JL2 Agreed. Brilliantly captured.

  • Any plant, whether designated a ‘weed’ or not, will change in its appearance over time, i.e. it will change its visual attribute states. Some of these changes will be positive – like the appearance of flowers in the spring; while others might be perceived as negative – like growing too big for its space during the summer and overwhelming adjacent plants, or going brown in autumn. Thus, ‘desirability’ of an attribute state is often (always?) context-dependent and is based upon a value judgement. Any weed is likely to be more attractive/acceptable at some times in the year than at others.

JL3 Agreed, but desirability reflects ‘intent’ as well as actual/possible state.

  • In the D&L conception a task seeks to change the (attribute) state of a domain object from a less desired state to a more desired state. As a performance metric, the closer the state change achieved by the IWS is to that desired, the higher the Tq achieved. However, in terms of the conception, the ‘desirability’ of the state being sought in the conduct of a task is arbitrary  – it could be ‘good’ or’ bad’, dependent on the point of view of the beholder.

JL4 Agreed but not keen on the use of the word ‘arbitrary’ here…..the desirer is a person or persons – the worker/gardener/commissioner/financier etc. ‘Beholder’ position not clear. Not just anyone, surely?

  • The human element (G in this case) undertaking a gardening task must perform an evaluative function, dynamically assessing progress towards the desired set of attribute states and adjusting their behaviour accordingly. This same evaluative function must presumably take place before work begins, in specifying the task in terms of the desired end state and an acceptable level of costs to G.

JL5 Agreed. Evaluation is an IWS process.

One might speculate that the mental representation of an object held by an individual may embody positive and negative connotations, established and potentially changing, over time. However, some of these connotations may be stronger than others as a consequence of education and/or particularly salient learning experiences. These stronger connotations might manifest in the way that the object is habitually perceived by that individual.

JL6 Agreed.

One might further speculate that G will have established a representation of the object weed over his multi-decade experience of gardening, during which weeds have been repeatedly reinforced as having attributes with negative connotations, such as rapid growth and spread;  often deep rooting and so hard to extractintrusive/invasive habits; as a group, usually visually unattractive in the garden; disruptive of intended plans for the space. In the context of these negative attributes, weeds detract from G’s positive experience of the garden and hence reduce Tq. Furthermore, their necessary removal, again given their negative attributes of proliferation and tenacity, are associated with high Uc.

JL7 Agreed. Could not have put it better……

In this context, it would be unsurprising that a changed approach to the management of the garden, involving a more nuanced connotation of weeds and their management, might give rise to conflicting experiences of this class of plants for G. Specifically, G’s conception of weeds as ‘bad’ has been established and reinforced over decades. Furthermore, G’s natural inclination to favour order and tidiness over (even managed) ‘disorder’ adds to the reinforcement of the view of weeds as ‘bad’.

JL8 You should have been a psychologist…….Ooops……

Implications:

  • Choice of vocabulary in specifying a task for analysis in IWS terms needs to be neutral, to avoid  inherent biases in connotation. Alternatively – unless it is self-evident – the particular connotation attached to an attribute whose state is to be changed needs to be made explicit.

JL9 Not sure about ‘neutral’ here. With respect to whom? Agreed about being made explicit.

A weed might be more neutrally defined as a  self-propagated native plant (SPNP?!) or perhaps just a ‘wild plant’. These could be specifically classified as being good, bad or conditionally welcome in a specific context.

JL10

Would be good to introduce the notion of contrast here. For example, flowers/non-flowers………garden/wild……domestic/wild. The contrast needs to serve desirability.

Note that there are several beneficial consequences of allowing an increased proportion of wild plants in the garden, such as more biodiversity and less detrimental impact on wildlife in the vicinity resulting from pesticide use, in addition to reduced physical Uc.

JL11 Exactement! Bien exprime.

  • Discussion point: Can performance be specified without some understanding of the mental representations held by the human component of an IWS?

JL12 Why not, if the desirer is the IWS commissioner?

On the basis of G’s experience, it looks as if there might be a requirement to make explicit a model of the domain as held by the human element of the IWS. Perhaps this could be quite crude yet could still assist in defining more explicitly/precisely the change in the state of the domain that would reflect high and low Tq for a given task.

JL13 This would be one way of doing it.

G might hold a (probably small) number of  internal representations [templates?] of the garden reflecting alternative potential [desired] formations of the space. These might be historical – how the garden was originally envisaged; how it was when JL assumed sole responsibility for maintaining it. Some might be more recent – the garden under a well implemented low maintenance regime.  Representations of plans for the garden at some future time would also be possible – for example  the TdeB accommodating a swimming pool or the garden…

JL14 Indeed.

The set of garden features might be the same in most of the representations, but representations of alternative formations would differ with respect to the constituent objects  and their desired attribute states. The default representation utilised when doing actual work in the garden would be that of the current desired formation of the garden.

JL15 Agreed.

The issue raised by G in relation to weeds might reflect interference between a historical representation and the current working (default) representation of the garden domain.

JL16 Agreed – see also JL8

  • Discussion point: How might the mental model of the human element in an IWS be represented and operationalised?

JL 17 Explicitly or implicitly?

I have some thoughts, but probably premature to consider at this point.

  • Discussion point: How to deal with G’s discomfort? Maybe changing the label ‘weed’ to ‘self-propagated native plant’ wouldn’t be sufficient….. Perhaps reframing garden tasks in terms of new ultimate goals for the garden, such as improved environmental sustainability and increased natural diversity would be more fruitful?

JL18 G’s Uc? The ultimate goal of the garden to date has been to contribute in a distinctive way to ‘Auzon village life’…… and in contrast to other gardens with the same goal.